JabRef References output

Author	Title	Year	Journal/Proceedings	Reftype	DOI/URL
Abdel-Hakim, A. E. & Farag, A. A.	CSIFT: A SIFT Descriptor with Color Invariant Characteristics [BibTeX]	2006	cvpr	inproceedings	DOI
BibTeX: @inproceedings{Abdel-Hakim2006, author = {Alaa E. Abdel-Hakim and Aly A. Farag}, title = {CSIFT: A SIFT Descriptor with Color Invariant Characteristics}, journal = {cvpr}, publisher = {IEEE Computer Society}, year = {2006}, volume = {02}, pages = {1978-1983}, doi = {http://doi.ieeecomputersociety.org/10.1109/CVPR.2006.95} }
Arentz, W. A. & Olstad, B.	Classifying offensive sites based on image content [Abstract] [BibTeX]	2004	Computer Vision and Image Understanding	article	URL
Abstract: This paper proposes a method for helping to identify adult web sites by using the image- content as means of detecting erotic material. The image content is classified by investigat- ing probable skin-regions, and extracting their feature vectors. These feature vectors are based on color-, texture-, contour-, placement-, and relative size-information for a given re- gion. The importance of the di?erent elements in the feature vector is determined by a ge- netic algorithm. For each picture, the algorithm gives the probability that a certain picture has erotic content. By mapping all the images in a web site, and running the image-based classifier on the whole collection, we were able to set up a histogram of images with regards to the log-likelihood of erotic content for each image. Hence giving a good overview of the web sites content and at the same time leaving room for errors in the image-based classifier. The algorithm proved to be quite successful in our tests where all 20 sites where classi- fied correctly. The image-based classifier is able to properly identify 89% of the evaluation images at an average processing speed of 11 images per second. Although this experiment focused on classifying adult web sites, small alterations to the system can be done, enabling classification of other kinds of images and web sites.
BibTeX: @article{Arentz2004, author = {Arentz, Will Archer and Olstad, Bjorn}, title = {Classifying offensive sites based on image content}, journal = {Computer Vision and Image Understanding}, year = {2004}, volume = {94}, number = {1-3}, pages = {295-310}, url = {http://citeseer.ifi.unizh.ch/661538.html} }
Balch, T., Khan, Z. & Veloso, M.	Automatically tracking and analyzing the behavior of live insect colonies [BibTeX]	2001	International Conference on Autonomous Agents	conference
BibTeX: @conference{Balch2001, author = {Balch, Tucker and Khan, Zia and Veloso, Manuela}, title = {Automatically tracking and analyzing the behavior of live insect colonies}, booktitle = {International Conference on Autonomous Agents}, year = {2001} }
Baumberg, A. & Hogg, D.	An Efficient Method for Contour Tracking using Active Shape Models [Abstract] [BibTeX]	1994		techreport
Abstract: There has been considerable research interest recently, in the areas of real time contour tracking and active shape models. This paper demonstrates how dynamic filtering can be used in combination with a flexible shape model to track an articulated non-rigid body in motion. The results show the method being used to track the silhouette of a walking pedestrian across a scene in real time. The active shape model used was generated automatically from real image data and incorporates variability in shape due to orientation as well as object flexibility. A Kalman filter is used to control spatial scale for feature search over successive frames and for contour refinement on an individual frame. Iterative refinement allows accurate contour localisation where feasible, although there is a trade-off between speed and accuracy. The shape model incorporates knowledge of the likely shape of the contour and speeds up tracking by reducing the number of system parameters. A further increase in speed is obtained by filtering the shape parameters independently.
BibTeX: @techreport{Baumberg1994, author = {Baumberg, A and Hogg, D}, title = {An Efficient Method for Contour Tracking using Active Shape Models}, year = {1994}, volume = {94.11}, number = {94.11}, pages = {--} }
Beck, D., Rees, G., Frith, C. & Lavie, N.	Neural correlates of change detection and change blindness [Abstract] [BibTeX]	2001	Nature Neuroscience	article	URL
Abstract: Functional magnetic resonance imaging (fMRI) of subjects attempting to detect a visual change occurring during a screen flicker was used to distinguish the neural correlates of change detection from those of change blindness. Change detection resulted in enhanced activity in the parietal and right dorsolateral prefrontal cortex as well as category-selective regions of the extrastriate visual cortex (for example, fusiform gyrus for changing faces). Although change blindness resulted in some extrastriate activity, the dorsal activations were clearly absent. These results demonstrate the importance of parietal and dorsolateral frontal activations for conscious detection of changes in properties coded in the ventral visual pathway, and thus suggest a key involvement of dorsal?ventral interactions in visual awareness.
BibTeX: @article{Beck2001, author = {Beck, Diane and Rees, Geraint and Frith, Christopher and Lavie, Nilli}, title = {Neural correlates of change detection and change blindness}, journal = {Nature Neuroscience}, year = {2001}, volume = {4}, pages = {645-450}, url = {http://www.nature.com/neuro/journal/v4/n6/full/nn0601_645.html} }
Belongie, S., Malik, J. & Puzicha, J.	Shape Matching and Object Recognition Using Shape Contexts [BibTeX]	2002	IEEE Trans. Pattern Anal. Mach. Intell.	article	DOI
BibTeX: @article{Belongie2002, author = {S. Belongie and J. Malik and J. Puzicha}, title = {Shape Matching and Object Recognition Using Shape Contexts}, journal = {IEEE Trans. Pattern Anal. Mach. Intell.}, publisher = {IEEE Computer Society}, year = {2002}, volume = {24}, number = {4}, pages = {509--522}, doi = {http://dx.doi.org/10.1109/34.993558} }
Benhamou, S.	How to reliably estimate the tortuosity of an animal's path: straightness, sinuosity, or fractal dimension? [Abstract] [BibTeX]	2004	J Theor Biol	article	DOI URL
Abstract: The tortuosity of an animal's path is a key parameter in orientation and searching behaviours. The tortuosity of an oriented path is inversely related to the efficiency of the orientation mechanism involved, the best mechanism being assumed to allow the animal to reach its goal along a straight line movement. The tortuosity of a random search path controls the local searching intensity, allowing the animal to adjust its search effort to the local profitability of the environment. This paper shows that (1) the efficiency of an oriented path can be reliably estimated by a straightness index computed as the ratio between the distance from the starting point to the goal and the path length travelled to reach the goal, but such a simple index, ranging between 0 and 1, cannot be applied to random search paths; (2) the tortuosity of a random search path, ranging between straight line movement and Brownian motion, can be reliably estimated by a sinuosity index which combines the mean cosine of changes of direction with the mean step length; and (3) in the current state of the art, the fractal analysis of animals' paths, which may appear as an alternative and promising way to measure the tortuosity of a random search path as a fractal dimension ranging between 1 (straight line movement) and 2 (Brownian motion), is only liable to generate artifactual results. This paper also provides some help for distinguishing between oriented and random search paths, and depicts a general, comprehensive framework for analysing individual animals' paths in a two-dimensional space.
BibTeX: @article{Benhamou2004, author = {Simon Benhamou}, title = {How to reliably estimate the tortuosity of an animal's path: straightness, sinuosity, or fractal dimension?}, journal = {J Theor Biol}, year = {2004}, volume = {229}, number = {2}, pages = {209--220}, url = {http://dx.doi.org/10.1016/j.jtbi.2004.03.016}, doi = {http://dx.doi.org/10.1016/j.jtbi.2004.03.016} }
Birchfield, S. & Tomasi, C.	Depth Discontinuities by Pixel-to-Pixel Stereo [Abstract] [BibTeX]	1998	International Conference on Computer Vision	inproceedings	URL
Abstract: An algorithm to detect depth discontinuities from astereo pair of images is presented. The algorithm matchesindividual pixels in corresponding scanline pairs while allowingoccluded pixels to remain unmatched, then propagatesthe information between scanlines by means of afast postprocessor. The algorithmhandles large untexturedregions, uses a measure of pixel dissimilarity that is insensitiveto image sampling, and prunes bad search nodes toincrease the speed of dynamic programming. The computationis relatively fast, taking about 1.5 microseconds perpixel per disparityon a workstation. Approximate disparitymaps and precise depth discontinuities (along both horizontaland vertical boundaries) are shown for five stereo imagescontaining textured, untextured, fronto-parallel, andslanted objects.
BibTeX: @inproceedings{Birchfield1998, author = {Birchfield, Stan and Tomasi, Carlo}, title = {Depth Discontinuities by Pixel-to-Pixel Stereo}, booktitle = {International Conference on Computer Vision}, year = {1998}, pages = {--}, url = {http://vision.stanford.edu} }
Black, M.	Explaining Optical Flow Events with Parameterized Spatio-temporal Models [Abstract] [BibTeX]	1999	International Conference on Computer Vision and Pattern Recognition	inproceedings
Abstract: A spatio-temporal representation for complex opticalflow events is developed that generalizes traditionalparameterized motion models (e.g. affine). These generativespatio-temporal models may be non-linear orstochastic and are event-specific in that they characterizea particular type of object motion (e.g. sittingor walking). Within a Bayesian framework weseek the appropriate model, phase, rate, spatial position,and scale to account for the image variation.The posterior distribution over this parameter spaceconditioned on image measurements is typically non-Gaussian. The distribution is represented using factoredsampling and is predicted and updated overtime using the Condensation algorithm. The resultingframework automatically detects, localizes, and recognizesmotion events.
BibTeX: @inproceedings{Black1999, author = {Black, Michael}, title = {Explaining Optical Flow Events with Parameterized Spatio-temporal Models}, booktitle = {International Conference on Computer Vision and Pattern Recognition}, year = {1999}, pages = {--} }
Bobick, A. & Ivanov, Y.	Action Recognition Using Probabilistic Parsing [Abstract] [BibTeX]	1998	International Conference on Computer Vision and Pattern Recognition	inproceedings
Abstract: A new approach to the recognition of temporal behaviors and activities is presented. The fundamental idea, inspired by work in speech recognition, is to divide the inference problem into two levels. The lower level is performed using standard independent probabilistic temporal event detectors such as hidden Markov models (HMMs) to propose candidate detections of low level temporal features. The outputs of these detectors provide the input stream for a stochastic context-free grammar parsing mechanism. The grammar and parser provide longer range temporal constraints, disambiguate uncertain low level detections, and allow the inclusion of a priori knowledge about the structure of temporal events in a given domain. To achieve such a system we provide techniques for generating a discrete symbol stream from continuous low level detectors, for enforcing temporal exclusion constraints during parsing, and for generating a control method for low level feature application based upon the current parsing state. We demonstrate the approach in several experiments using both visual and other sensing data.
BibTeX: @inproceedings{Bobick1998, author = {Bobick, Aaron and Ivanov, Yuri}, title = {Action Recognition Using Probabilistic Parsing}, booktitle = {International Conference on Computer Vision and Pattern Recognition}, publisher = {IEEE Computer Society}, year = {1998}, pages = {196--202} }
Bosson, A., Cawley, G. C., Chan, Y. & Harvey, R.	Nonretrieval: blocking pornographic images [Abstract] [BibTeX]	2002	International Conference on Image and Video Retrieval	conference	URL
Abstract: We extend earlier work on detecting pornographic images. Our focus is on the classification stage and we give new results for a variety of classical and modern classifiers. We find the artificial neural network o?ers a statistically significant improvement. In all cases the error rate is too high unless deployed sensitively so we show how such a system may be built into a commercial environment.
BibTeX: @conference{Bosson2002, author = {Bosson, Alison and Cawley, Gavin C. and Chan, Yi and Harvey, Richard}, title = {Nonretrieval: blocking pornographic images}, booktitle = {International Conference on Image and Video Retrieval}, year = {2002}, url = {http://citeseer.ist.psu.edu/bosson02nonretrieval.html} }
Boutin, F. & Hascoet, M.	Cluster Validity Indices for Graph Partitioning [BibTeX]	2004	International Conference on Information Visualisation	conference	DOI
BibTeX: @conference{Boutin2004, author = {Boutin, Francois and Hascoet, Mountaz}, title = {Cluster Validity Indices for Graph Partitioning}, booktitle = {International Conference on Information Visualisation}, publisher = {IEEE Computer Society}, year = {2004}, pages = {376--381}, doi = {http://dx.doi.org/10.1109/IV.2004.36} }
Brand, J. & Mason, J. S.	A Comparative Assessment of Three Approaches to Pixel-level Human Skin-Detection [Abstract] [BibTeX]	2000	International Conference on Pattern Recognition	conference
Abstract: This paper assesses the merits of three different approaches to pixel-level human skin detection. The basis for the three approaches has been reported recently in the literature. The first two approaches [1, 2] use simple ratios and color space transforms respectively, whereas the third is a numerically efficient approach based on a 3-D RGB probability map, first implemented by Re-hg [3]. The Bayesian probabilities are made possible to compute only with the availability of a large appropriately labeled database. Over 12,000 images from the Compaq skin and non-skin databases [4] are used to quantitatively assess the three approaches. Thresholds are determined empirically to detect 95% of all skin-associate d pixels and assessment is then made in terms of the percentage of non-skin pixels incorrectly accepted. The lowest of these false acceptance rates is found to be about 20% given by the 3-D probability map.
BibTeX: @conference{Brand2000, author = {Brand, Jason and Mason, John S.}, title = {A Comparative Assessment of Three Approaches to Pixel-level Human Skin-Detection}, booktitle = {International Conference on Pattern Recognition}, year = {2000} }
Brand, M.	Coupled hidden Markov models for modeling interacting processes [Abstract] [BibTeX]	1996	Neural Computation	article
Abstract: We present methods for coupling hidden Markov models (HMMs) to model systems of multiple interacting processes. The resulting models have multiple state variables that are temporally coupled via matrices of conditional probabilities. We introduce a deterministic O(T(CN)2) approximation for maximum a posterior (MAP) state estimation which enables fast classification and parameter estimation via expectation maximization. An ``N-heads'' dynamic programming algorithm samples from the highest probability paths through a compact state trellis, minimizing an upper bound on the cross entropy with the full (combinatoric) dynamic programming problem. The complexity is O(T(CN)2) for C chains of N states apiece observing T data points, compared with O(TN2C) for naive (Cartesian product), exact (state clustering), and stochastic (Monte Carlo) methods applied to the same inference problem. In several experiments examining training time, model likelihoods, classification accuracy, and robustness to initial conditions, coupled HMMs compared favorably with conventional HMMs and with energy-based approaches to coupled inference chains. We demonstrate and compare these algorithms on synthetic and real data, including interpretation of video.
BibTeX: @article{Brand1996, author = {Brand, Matthew}, title = {Coupled hidden Markov models for modeling interacting processes}, journal = {Neural Computation}, year = {1996}, number = {405} }
Bras Silva, H.	A partitional clustering algorithm validated by a clustering tendency index based on graph theory [Abstract] [BibTeX]	2006	Pattern Recognition	article
Abstract: Applying graph theory to clustering, we propose a partitional clustering method and a clustering tendency index. No initial assumptions about the data set are requested by the method. The number of clusters and the partition that best fits the data set, are selected according to the optimal clustering tendency index value.
BibTeX: @article{Bras2006, author = {Bras Silva, Helena}, title = {A partitional clustering algorithm validated by a clustering tendency index based on graph theory}, journal = {Pattern Recognition}, year = {2006} }
Broggi, A. & Fascioli, A.	Artificial Vision in Extreme Environments for Snowcat Tracks Detection [Abstract] [BibTeX]	2002	IEEE Transactions on Intelligent Transportation Systems	article
Abstract: This paper describes the image processing techniquesdesigned to localize the tracks of snowcats for the automation oftransportation of goods and people during the Italian scientificmissions in Antarctica. The final goal is to enable a snowcat toautomatically follow the preceding one in a train-like fashion. Acamera is used to acquire images of the scene; the image sequenceis analyzed by a computer vision system which identifies the tracksand produces a high level description of the scene. This result isthen forwarded to a further software module in charge of the controlof the snowcat movement. A further optional representation,in which markers highlighting the tracks are superimposed ontothe acquired image, is transmitted to a human supervisor locatedoff board. This system has been tested in the Italian test site andwas under testing in the South Pole during the early 2002 Italianscientific mission. The paper also briefly describes an alternativesolution based on an evolutionary approach.
BibTeX: @article{Broggi2002, author = {Broggi, Alberto and Fascioli, Alessandra}, title = {Artificial Vision in Extreme Environments for Snowcat Tracks Detection}, journal = {IEEE Transactions on Intelligent Transportation Systems}, year = {2002}, volume = {3}, number = {3} }
Brown, D., Craw, I. & Lewthwaite, J.	A SOM Based Approach to Skin Detection with Application in Real Time Systems [Abstract] [BibTeX]	2001	British Machine Vision Conference	conference	URL
Abstract: A large body of human image processing techniques use skin detection as a first primitive for subsequent feature extraction. Well established methods of colour modelling, such as histograms and Gaussian mixture models have enabled the construction of suitably accurate skin detectors. However such techniquesare notideal forusein adaptiverealtimeenvironments. We describe methods of skin detection using a Self-Organising Map or SOM, and show performance comparable (94% accuracy) to conventional techniques. We also introduce the AXEON Learning Processor as the basis for a hardware skin detector, and outline the potential benefits of using this systeminademandingenvironment,suchasfilteringInternettraffic,towhich conventionaltechniquesare notbest suited.
BibTeX: @conference{Brown2001, author = {Brown, David and Craw, Ian and Lewthwaite, Julian}, title = {A SOM Based Approach to Skin Detection with Application in Real Time Systems}, booktitle = {British Machine Vision Conference}, year = {2001}, url = {http://citeseer.ist.psu.edu/brown01som.html} }
Bruce, J., Balch, T. & Veloso, M.	Fast and Inexpensive Color Image Segmentation for Interactive Robots [Abstract] [BibTeX]	2000	International Conference on Intelligent Robots and Systems	inproceedings	URL
Abstract: Vision systems employing region segmentation by colorare crucial in real-time mobile robot applications, such asRoboCup[1], or other domains where interaction with humansor a dynamic world is required. Traditionally, systemsemploying real-time color-based segmentation are eitherimplemented in hardware, or as very specific softwaresystems that take advantage of domain knowledge to attainthe necessary efficiency. However, we have found that withcareful attention to algorithm efficiency, fast color imagesegmentation can be accomplished using commodity imagecapture and CPU hardware. Our paper describes asystem capable of tracking several hundred regions of upto 32 colors at 30 Hertz on general purpose commodityhardware. The software system is composed of four mainparts; a novel implementation of a threshold classifier, amerging system to form regions through connected components,a separation and sorting system that gathers variousregion features, and a top down merging heuristic toapproximate perceptual grouping. A key to the efficiencyof our approach is a new method for accomplishing colorspace thresholding that enables a pixel to be classified intoone or more of up to 32 colors using only two logical ANDoperations. A naive approach could require up to 192 comparisonsfor the same classification. The algorithms andrepresentations are described, as well as descriptions ofthree applications in which it has been used.
BibTeX: @inproceedings{Bruce2000, author = {Bruce, James and Balch, Tucker and Veloso, Manuela}, title = {Fast and Inexpensive Color Image Segmentation for Interactive Robots}, booktitle = {International Conference on Intelligent Robots and Systems}, year = {2000}, pages = {--}, url = {www.cs.cmu.edu/~trb/papers/wirevision00.pdf} }
Bui, H., Venkatesh, S. & West, G.	Tracking and surveillance in wide-area spatial environments using the abstract Hidden Markov Model [Abstract] [BibTeX]	2001	International Journal of Pattern Recognition and Artificial Intelligence	article
Abstract: In this paper, we consider the problem of tracking an object and predicting the object's future trajectory in a wide-area environment, with complex spatial layout and the use of multiple sensors/cameras. To solve this problem, there is a need for representing the dynamic and noisy data in the tracking tasks, and dealing with them at different levels of detail. We employ the Abstract Hidden Markov Models (AHMM), an extension of the well-known Hidden Markov Model (HMM) and a special type of Dynamic Probabilistic Network (DPN), as our underlying representation framework. The AHMM allows us to explicitly encode the hierarchy of connected spatial locations, making it scalable to the size of the environment being modeled. We describe an application for tracking human movement in an office-like spatial layout where the AHMM is used to track and predict the evolution of object trajectories at different levels of detail
BibTeX: @article{Bui2001, author = {Bui, H.H. and Venkatesh, S. and West, G.}, title = {Tracking and surveillance in wide-area spatial environments using the abstract Hidden Markov Model}, journal = {International Journal of Pattern Recognition and Artificial Intelligence}, year = {2001}, volume = {15}, number = {1}, pages = {177--195} }
Burke, R.	The Third Wave of Marketing Intelligence [BibTeX]	2005	Retailing in the 21st Century	incollection
BibTeX: @incollection{Burke2005, author = {Burke, Raymond}, title = {The Third Wave of Marketing Intelligence}, booktitle = {Retailing in the 21st Century}, publisher = {Springer-Verlag}, year = {2005} }
Buxton, H.	Generative Models for Learning and Understanding Dynamic Scene Activity [Abstract] [BibTeX]	2002	International Workshop on Generative-Model-Based Vision	inproceedings	URL
Abstract: We are entering an era of more intelligent cognitive vision systems.Such systems can analyse activity in dynamic scenes to computeconceptual descriptions from motion trajectories of moving people andthe objects they interact with. Here we review progress in the development of flexible, generative models that can explain visual input as a combination of hidden variables and can adapt to new typesof input. Such models are particularly appropriate for the tasksposed by cognitive vision as they incorporate learning as well as having sufficient structure to represent a general class of problems.In addition, generative models explain all aspects of the inputrather than attempting to ignore irrelevant sources of variation asin exemplar-based learning. Applications of these models in visualinteraction for education, smart rooms and cars, as well assurveillance systems is also briefly reviewed.
BibTeX: @inproceedings{Buxton2002, author = {Buxton, H}, title = {Generative Models for Learning and Understanding Dynamic Scene Activity}, booktitle = {International Workshop on Generative-Model-Based Vision}, year = {2002}, pages = {--}, url = {http://citeseer.ist.psu.edu/buxton02generative.html} }
Buzan, D., Sclaroff, S. & Kollios, G.	Extraction and Clustering of Motion Trajectories in Video [Abstract] [BibTeX]	2004	International Conference on Pattern Recognition	inproceedings	DOI
Abstract: A system is described that tracks moving objects in a video dataset so as to extract a representation of the objects’ 3D trajectories. The system then finds hierarchical clusters of similar trajectories in the video dataset. Objects’ motion trajectories are extracted via an EKF formulation that provides each object’s 3D trajectory up to a constant factor. To increase accuracy when occlusions occur, multiple tracking hypotheses are followed. For trajectory-based clustering and retrieval, a modified version of edit distance, called longest common subsequence is employed. Similarities are computed between projections of trajectories on coordinate axes. Trajectories are grouped based, using an agglomerative clustering algorithm. To check the validity of the approach, experiments using real data were performed.
BibTeX: @inproceedings{Buzan2004, author = {Buzan, Dan and Sclaroff, Stan and Kollios, George}, title = {Extraction and Clustering of Motion Trajectories in Video}, booktitle = {International Conference on Pattern Recognition}, publisher = {IEEE Computer Society}, year = {2004}, pages = {521--524}, doi = {http://dx.doi.org/10.1109/ICPR.2004.356} }
Caillette, F. & Howard, T.	Real-Time Markerless Human Body Tracking with Multi-View 3-D Voxel Reconstruction [Abstract] [BibTeX]	2004	British Machine Vision Conference	inproceedings
Abstract: We present an approach to full human-body tracking, using markerless multiviewimages as input, performing acquisition, reconstruction and tracking inreal-time on a single PC. Our approach employs a hierarchical visual-hull algorithmwhich segments only the most interesting regions of the images andincludes colour information. The tracking step uses blobs attached to a kinematicmodel to recover joint angles in an expectation-maximization framework.We demonstrate the robustness of the approach on video sequences ofvarious body configurations in an unaugmented office environment. We alsoshow that tracking challenging poses with self-occlusions is possible withoutthe processing cost of stochastic sampling schemes.
BibTeX: @inproceedings{Caillette2004, author = {Caillette, Fabrice and Howard, Toby}, title = {Real-Time Markerless Human Body Tracking with Multi-View 3-D Voxel Reconstruction}, booktitle = {British Machine Vision Conference}, year = {2004}, pages = {--} }
CAVIAR	IST 37540 [BibTeX]	2001		misc	URL
BibTeX: @misc{CAVIAR2001, author = {CAVIAR}, title = {IST 37540}, year = {2001}, url = {http://groups.inf.ed.ac.uk/vision/CAVIAR/CAVIARDATA1/} }
Chai, D. & Ngan, K.	Locating facial region of a head-and-shoulders color image [BibTeX]	1998	International Conference on Automatic Face and Gesture Recognition	conference
BibTeX: @conference{Chai1998, author = {Chai, Douglas and Ngan, King}, title = {Locating facial region of a head-and-shoulders color image}, booktitle = {International Conference on Automatic Face and Gesture Recognition}, year = {1998} }
Cheng, S. Y., Park, S. & Trivedi, M. M.	Multiperspective Thermal IR and Video Arrays for 3D Body Tracking and Driver Activity Analysis [BibTeX]	2005	CVPR '05: Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) - Workshops	inproceedings	DOI
BibTeX: @inproceedings{Cheng2005, author = {Shinko Y. Cheng and Sangho Park and Mohan M. Trivedi}, title = {Multiperspective Thermal IR and Video Arrays for 3D Body Tracking and Driver Activity Analysis}, booktitle = {CVPR '05: Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) - Workshops}, publisher = {IEEE Computer Society}, year = {2005}, pages = {3}, doi = {http://dx.doi.org/10.1109/CVPR.2005.500} }
Cheng, Y.	Mean Shift, Mode Seeking, and Clustering [Abstract] [BibTeX]	1995	IEEE Transactions on Pattern Analysis and Machine Intelligence	article	DOI
Abstract: Mean shift, a simple iterative procedure that shifts each data point to the average of data points in its neighborhood, is generalized and analyzed in this paper. This generalization makes some k-means like clustering algorithms its special cases. It is shown that mean shift is a mode-seeking process on a surface constructed with a ¿shadow¿ kernel. For Gaussian kernels, mean shift is a gradient mapping. Convergence is studied for mean shift iterations. Cluster analysis is treated as a deterministic problem of finding a fixed point of mean shift that characterizes the data. Applications in clustering and Hough transform are demonstrated. Mean shift is also considered as an evolutionary strategy that performs multistart global optimization.
BibTeX: @article{Cheng1995, author = {Cheng, Yizong}, title = {Mean Shift, Mode Seeking, and Clustering}, journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence}, publisher = {IEEE Computer Society}, year = {1995}, volume = {17}, number = {8}, pages = {790--799}, doi = {http://dx.doi.org/10.1109/34.400568} }
Chib, S. & Greenberg, E.	Understanding the Metropolis-Hastings Algorithm [Abstract] [BibTeX]	1995	The American Statistician	article	URL
Abstract: We provide a detailed, introductory exposition of the Metropolis-Hastings algorithm, a powerful Markov chain method to simulate multivariate distributions. A simple, intuitive derivation of this method is given along with guidance on implementation. Also discussed are two applications of the algorithm, one for implementing acceptance-rejection sampling when a blanketing function is not available and the other for implementing the algorithm with block-at-a-time scans. In the latter situation, many different algorithms, including the Gibbs sampler, are shown to be special cases of the Metropolis-Hastings algorithm. The methods are illustrated with examples.
BibTeX: @article{Chib1995, author = {Chib, Siddhartha and Greenberg, Edward }, title = {Understanding the Metropolis-Hastings Algorithm}, journal = {The American Statistician}, year = {1995}, volume = {49}, number = {4}, pages = {327--335}, url = {http://links.jstor.org/sici?sici=0003-1305%28199511%2949%3A4%3C327%3AUTMA%3E2.0.CO%3B2-2} }
Chilgunde, A., Kumar, P., Ranganath, S. & WeiMin, H.	Multi-Camera Target Tracking in Blind Regions of Cameras with Non-overlapping Fields of View [Abstract] [BibTeX]	2004	British Machine Vision Conference	inproceedings
Abstract: In this paper, we propose a real time system for tracking targets acrossblind regions of multiple cameras with non-overlapping fields of views (FOVs)using camera topology, and targets' motion and shape information. Kalmanfilters are used to robustly track each target's shape and motion in each cameraview and the common ground plane view composed of all camera views.The target's track in the blind region between cameras is obtained usingKalman filter predictions. For multi-camera correspondence matching wecompute the Gaussian distributions of the tracking parameters across camerasfor the target motion and position in the ground plane view. Matching oftargets across camera views uses a graph based track initialization scheme,which accumulates information from occurrences of target in several consecutiveframes of the video. Probabilistic matching is carried out by using thetrack parameters for new tracks obtained from the graph in a camera viewwith the parameters of the terminated tracks learnt by Kalman filters in theother camera views and ground plane view. We obtain 85% accuracy forcorresponding matching while tracking vehicles observed from two camerasmonitoring a highway.
BibTeX: @inproceedings{Chilgunde2004, author = {Chilgunde, Amit and Kumar, Pankaj and Ranganath, Surendra and WeiMin, Huang}, title = {Multi-Camera Target Tracking in Blind Regions of Cameras with Non-overlapping Fields of View}, booktitle = {British Machine Vision Conference}, year = {2004}, pages = {--} }
Cipolla, R., Drummond, T. & Robertson, D.	Camera Calibration from Vanishing Points in Images of Architectural Scenes [Abstract] [BibTeX]	1999	British Machine Vision Conference	inproceedings
Abstract: We address the problem of recovering 3D models from uncalibratedimages of architectural scenes. We propose a simple, geometrically in-tuitive method which exploits the strong rigidity constraints of paral-lelism and orthogonality present in indoor and outdoor architecturalscenes. We present a novel algorithm that uses these simple constraintsto recover the projection matrices for each viewpoint and relate ourmethod to the algorithm of Caprile and Torre [2].The projection matrices are used to recover partial 3D models ofthe scene and these can be used to visualise new viewpoints. Ourapproach does not need any a priori information about the camerasbeing used. A working system called PhotoBuilder has been designedand implemented to allow a user to interactively build a VRML modelof a building from uncalibrated images from arbitrary viewpoints [3, 4].
BibTeX: @inproceedings{Cipolla1999, author = {Cipolla, R and Drummond, Tom and Robertson, D}, title = {Camera Calibration from Vanishing Points in Images of Architectural Scenes}, booktitle = {British Machine Vision Conference}, year = {1999}, volume = {II}, pages = {382-392} }
Collins, R.	Mean-shift Blob Tracking through Scale Space [Abstract] [BibTeX]	2003	International Conference on Computer Vision and Pattern Recognition	inproceedings
Abstract: The mean-shift algorithm is an efficient technique for tracking 2D blobs through an image. Although the scale of the mean-shift kernel is a crucial parameter, there is presently no clean mechanism for choosing or updating scale while tracking blobs that are changing in size. We adapt Lindeberg’s theory of feature scale selection based on local maxima of differential scale-space filters to the problem of selecting kernel scale for mean-shift blob tracking. We show that a difference of Gaussian (DOG) mean-shift kernel enables efficient tracking of blobs through scale space. Using this kernel requires generalizing the mean-shift algorithm to handle images that contain negative sample weights.
BibTeX: @inproceedings{Collins2003, author = {Collins,Robert}, title = {Mean-shift Blob Tracking through Scale Space}, booktitle = {International Conference on Computer Vision and Pattern Recognition}, year = {2003} }
Collins, R., Lipton, A., Kanade, T., Fujiyoshi, H., Duggins, D., Tsin, Y., Tolliver, D., Enomoto, N., Hasegawa, O., Burt, P. & Wixson, L.	A System for Video Surveillance and Monitoring [BibTeX]	2000		techreport
BibTeX: @techreport{Collins2000, author = {Collins,Robert and Lipton, Alan and Kanade,Takeo and Fujiyoshi, Hironobu and Duggins, David and Tsin, Yanghai and Tolliver, David and Enomoto, Nobuyoshi and Hasegawa, Osamu and Burt, Peter and Wixson, Lambert}, title = {A System for Video Surveillance and Monitoring}, year = {2000}, number = {CMU-RI-TR-00-12} }
Collins, R. & Weiss, R.	Vanishing Point Calculation as a Statistical Inference on the Unit Sphere [Abstract] [BibTeX]	1990	International Conference on Computer Vision	inproceedings
Abstract: An examination is made of vanishing point calculation as a statistical estimation problem. It is assumed that image line segments have been previously clustered into groups of convergent lines. For each group, the vanishing point location is estimated as the polar axis of an equatorial distribution on the unit sphere, and the statistical error of the estimate is determined. The sensitivity of the estimates to the number of lines in a convergent cluster is studied.
BibTeX: @inproceedings{Collins1990, author = {Collins, Robert and Weiss, R}, title = {Vanishing Point Calculation as a Statistical Inference on the Unit Sphere}, booktitle = {International Conference on Computer Vision}, year = {1990} }
Comaniciu, D., Meer, P. & Ramesh, V.	Kernel-based object tracking [Abstract] [BibTeX]	2003	IEEE Transactions on Pattern Analysis and Machine Intelligence	article	URL
Abstract: A new approach toward target representation and localization, the central component in visual tracking of nonrigid objects, is proposed. The feature histogram-based target representations are regularized by spatial masking with an isotropic kernel. The masking induces spatially-smooth similarity functions suitable for gradient-based optimization, hence, the target localization problem can be formulated using the basin of attraction of the local maxima. We employ a metric derived from the Bhattacharyya coefficient as similarity measure, and use the mean shift procedure to perform the optimization. In the presented tracking examples, the new method successfully coped with camera motion, partial occlusions, clutter, and target scale variations. Integration with motion filters and data association techniques is also discussed. We describe only a few of the potential applications: exploitation of background information, Kalman tracking using motion models, and face tracking.
BibTeX: @article{Comaniciu2003, author = {Comaniciu, Dorin and Meer, Peter and Ramesh, Visvanathan}, title = {Kernel-based object tracking}, journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence}, year = {2003}, volume = {25}, number = {5}, pages = {564--575}, url = {http://www.caip.rutgers.edu/%7Ecomanici/} }
Connaire, C. O., O'Connor, N. & Smeaton., A. F.	Thermo-Visual Feature Fusion for Object Tracking Using Multiple Spatiogram Trackers [BibTeX]	2007	Machine Vision and Applications	article	DOI
BibTeX: @article{Connaire2007, author = {Ciaran O Connaire and Noel O'Connor and Alan F. Smeaton.}, title = {Thermo-Visual Feature Fusion for Object Tracking Using Multiple Spatiogram Trackers}, journal = {Machine Vision and Applications}, publisher = {Springer}, year = {2007}, pages = {1--12}, doi = {http://dx.doi.org/10.1007/s00138-007-0078-y} }
Criminisi, A., Reid, I. & Zisserman, A.	A plane measuring device [Abstract] [BibTeX]	1999	Image and Vision Computing	article	URL
Abstract: A requirement of a visual measurement device is that both measurements and their uncertainties can be determined. This paper develops an uncertainty analysis which includes both the errors in image localization and the uncertainty in the imaging transformation. The matrix representing the imaging transformation is estimated from image-to-world point correspondences. A,general expression is derived for the covariance of this matrix. This expression is valid if the matrix is over determined and also if the minimum number of correspondences are used. A bound on the errors of the first order approximations involved is also derived. Armed with this covariance result the uncertainty of any measurement can be predicted, and furthermore the distribution of correspondences can be chosen to achieve a particular bound on the uncertainty. Examples are given of measurements such as distance and parallelism for several applications. These include indoor scenes and architectural measurements. (C) 1999 Elsevier Science B.V. All rights reserved
BibTeX: @article{Criminisi1999, author = {Criminisi, Antonio and Reid, I. and Zisserman, A.}, title = {A plane measuring device}, journal = {Image and Vision Computing}, year = {1999}, volume = {17(8)}, number = {8}, pages = {625--634}, url = {http://www.robots.ox.ac.uk/~vgg/presentations/bmvc97/criminispaper/planedev.html} }
Criminisi, A., Reid, I. & Zisserman, A.	Single view metrology [Abstract] [BibTeX]	1999	International Journal of Computer Vision	article	URL
Abstract: We describe how 3D affine measurements may be computed from a single perspective view of a scene given only minimal geometric information determined from the image. This minimal information is typically the vanishing line of a reference plane, and a vanishing point for a direction not parallel to the plane. It is shown that affine scene structure may then be determined from the image, without knowledge of the camera's internal calibration (e.g. focal length), nor of the explicit relation between camera and world (pose). In particular, we show how to (i) compute the distance between planes parallel to the reference plane (up to a common scale factor); (ii) compute area and length ratios on any plane parallel to the reference plane; (iii) determine the camera's location. Simple geometric derivations are given for these results. We also develop an algebraic representation which unifies the three types of measurement and, amongst other advantages, permits a first order error propagation analysis to be performed, associating an uncertainty with each measurement. We demonstrate the technique for a variety of applications, including height measurements in forensic images and 3D graphical modelling from single images
BibTeX: @article{Criminisi1999a, author = {Criminisi, Antonio and Reid, I. and Zisserman, A.}, title = {Single view metrology}, journal = {International Journal of Computer Vision}, year = {1999}, volume = {40}, number = {2}, pages = {123--148}, url = {ISI:000166197700002} }
Criminisi, A., Zisserman, A., Van Gool, L. & Bramble, S.	A New Approach to Obtain Height Measurements from Video [Abstract] [BibTeX]	1998	Proceedings of The International Society for Optical Engineering	inproceedings	URL
Abstract: In this paper a new measurement algorithm is presented which generates height measurements and their associated errors from a single known physical measurement in an image. The method draws on results from projective geometry and computer vision. A height measurement is obtained from each frame of the video. A `stereo like' correspondence between images is not required. Nor is any explicit camera calibration. The accuracy of the algorithm is demonstrated by a number of examples when ground truth is known. Finally, the height measurements and their variation are described for a person in motion. We draw attention to the uncertainty in heights associated with humans in motion, and the limitations of using this description for identification.
BibTeX: @inproceedings{Criminisi1998, author = {Criminisi, Antonio and Zisserman, Andrew and Van Gool, Luc and Bramble, Simon}, title = {A New Approach to Obtain Height Measurements from Video}, booktitle = {Proceedings of The International Society for Optical Engineering}, year = {1998}, pages = {227-238}, url = {http://www.robots.ox.ac.uk/~vgg/presentations/spie98/criminis/p0.html} }
Cucchiara, R., Grana, C., Piccardi, M. & Prati, A.	Detecting Moving Objects, Ghosts, and Shadows in Video Streams [BibTeX]	2003	IEEE Transactions on Pattern Analysis and Machine Intelligence	article	DOI
BibTeX: @article{Cucchiara2003, author = {Rita Cucchiara and Costantino Grana and Massimo Piccardi and Andrea Prati}, title = {Detecting Moving Objects, Ghosts, and Shadows in Video Streams}, journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence}, publisher = {IEEE Computer Society}, year = {2003}, volume = {25}, number = {10}, pages = {1337-1342}, doi = {http://doi.ieeecomputersociety.org/10.1109/TPAMI.2003.1233909} }
Cupillard, F., Brémond, F. & Thonnat, M.	GROUP BEHAVIOR RECOGNITION WITH MULTIPLE CAMERAS [BibTeX]	2002	wacv	article	DOI
BibTeX: @article{Cupillard2002, author = {Frédéric Cupillard and François Brémond and Monique Thonnat}, title = {GROUP BEHAVIOR RECOGNITION WITH MULTIPLE CAMERAS}, journal = {wacv}, publisher = {IEEE Computer Society}, year = {2002}, volume = {0}, pages = {177}, doi = {http://doi.ieeecomputersociety.org/10.1109/ACV.2002.1182178} }
Dai, C., Zheng, Y. & Li, X.	Layered Representation for Pedestrian Detection and Tracking in Infrared Imagery [BibTeX]	2005	CVPR '05: Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) - Workshops	inproceedings	DOI
BibTeX: @inproceedings{Dai2005, author = {Congxia Dai and Yunfei Zheng and Xin Li}, title = {Layered Representation for Pedestrian Detection and Tracking in Infrared Imagery}, booktitle = {CVPR '05: Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) - Workshops}, publisher = {IEEE Computer Society}, year = {2005}, pages = {13}, doi = {http://dx.doi.org/10.1109/CVPR.2005.483} }
Davies, D. & Bouldin, D.	A Cluster Separation Measure [BibTeX]	1979	International Journal of Pattern Recognition and Artificial Intelligence	article
BibTeX: @article{Davies1979, author = {D.L. Davies and D.W. Bouldin}, title = {A Cluster Separation Measure}, journal = {International Journal of Pattern Recognition and Artificial Intelligence}, year = {1979}, volume = {1 (2)}, pages = {224-227} }
Davis, J. & Sharma, V.	Fusion-Based Background-Subtraction using Contour Saliency [Abstract] [BibTeX]	2005	International Conference on Computer Vision and Pattern Recognition	conference	URL
Abstract: We present a new contour-based background-subtraction technique using thermal and visible imagery for persistent object detection in urban settings. Statistical backgroundsubtraction in the thermal domain is used to identify the initial regions-of-interest. Color and intensity information are used within these areas to obtain the corresponding regionsof- interest in the visible domain. Within each region, input and background gradient information are combined to form a Contour Saliency Map. The binary contour fragments, obtained from corresponding Contour Saliency Maps, are then combined. An A path-constrained search along watershed boundaries is used to complete and close any broken contour segments. Lastly, the contour image is flood- filled to produce silhouettes. Results of our approach are presented and compared against manually segmented data.
BibTeX: @conference{Davis2005, author = {Davis, James and Sharma, Vinay}, title = {Fusion-Based Background-Subtraction using Contour Saliency}, booktitle = {International Conference on Computer Vision and Pattern Recognition}, year = {2005}, pages = {III: 11-11}, url = {http://www.cse.ohio-state.edu/~jwdavis/} }
Dee, H. & Hogg, D.	Detecting inexplicable behaviour [Abstract] [BibTeX]	2004	British Machine Vision Conference	inproceedings
Abstract: This paper presents a novel approach to the detection of unusual or interestingevents in videos involving certain types of intentional behaviour, suchas pedestrian scenes. The approach is not based upon a statistical measureof typicality, but upon building an understanding of the way people navigatetowards a goal. The activity of agents moving around within the scene isevaluated based upon whether the behaviour in question is consistent witha simple model of goal-directed behaviour and a model of those goals andobstacles known to be in the scene. The advantages of such an approach aremultiple: it handles the presence of movable obstacles (for example, parkedcars) with ease; trajectories which have never before been presented to thesystem can be classied as explicable; and the technique as a whole has aprima facie psychological plausibility. A system based upon these principlesis demonstrated in two scenes: a car-park, and in a foyer scenario1.
BibTeX: @inproceedings{Dee2004, author = {Dee, Hannah and Hogg, David}, title = {Detecting inexplicable behaviour}, booktitle = {British Machine Vision Conference}, year = {2004}, pages = {477–486} }
Deutscher, J., North, B., Bascle, B. & Blake	Tracking through singularities and discontinuities by random sampling [Abstract] [BibTeX]	1999	International Conference on Computer Vision	inproceedings	URL
Abstract: Some issues in markerless tracking of human body motionare addressed. Extended Kalman filters have commonly beenapplied to kinematic variables, to combine predictions consistentwith plausible motion, with the incoming stream of visualmeasurements. Kalman filtering is applicable only whenthe underlying distribution is approximately Gaussian. Often,this assumption proves remarkably robust.There are two pervasive circumstances under which theGaussianity assumption can break down. The first is kinematicsingularity, and the second is at joint endstops. Failureof Kalman filtering under these circumstance is illustrated.The non-Gaussian nature of the distributions isdemonstrated experimentally by means of Monte-Carlo simulation.Random simulation - particle filtering or Condensation-proves to provide a robust alternative algorithm fortracking that can also deal with these difficult conditions.
BibTeX: @inproceedings{Deutscher1999, author = {Deutscher, J and North, B and Bascle, B and Blake}, title = {Tracking through singularities and discontinuities by random sampling}, booktitle = {International Conference on Computer Vision}, publisher = {IEEE Computer Society}, year = {1999}, pages = {1144}, url = {www.robots.ox.ac.uk/~vdg} }
Doucet, A., de Freitas, N. & Gordon, N.	Sequential Monte Carlo Methods in Practice [BibTeX]	2001		book
BibTeX: @book{Doucet2001, author = {Doucet, Arnaud and de Freitas, Nando and Gordon, Neil}, title = {Sequential Monte Carlo Methods in Practice}, publisher = {Springer-Verlag}, year = {2001} }
Efros, A., Berg, A., Mori, G. & Malik, J.	Recognizing Action at a Distance [Abstract] [BibTeX]	2003	International Conference on Computer Vision	conference
Abstract: Our goal is to recognize human actions at a distance, at resolutions where a whole person may be, say, 30 pixels tall. We introduce a novel motion descriptor based on optical flow measurements in a spatio-temporal volume for each stabilized human figure, and an associated similarity measure to be used in a nearest-neighbor framework. Making use of noisy optical flow measurements is the key challenge, which is addressed by treating optical flow not as precise pixel displacements, but rather as a spatial pattern of noisy measurements which are carefully smoothed and aggregated to form our spatio-temporal motion descriptor. To classify the action being performed by a human figure in a query sequence, we retrieve nearest neighbor(s) from a database of stored, annotated video sequences. We can also use these retrieved exemplars to transfer 2D/3D skeletons onto the figures in the query sequence, as well as two forms of data-based action synthesis ``Do as I Do'' and ``Do as I Say''. Results are demonstrated on ballet, tennis as well as football datasets.
BibTeX: @conference{Efros2003, author = {Efros, Alexei and Berg, Alexander and Mori, Greg and Malik, Jitendra}, title = {Recognizing Action at a Distance}, booktitle = {International Conference on Computer Vision}, year = {2003} }
Ekman, P.	Basic Emotions [BibTeX]	1999		inbook
BibTeX: @inbook{Ekman1999a, author = {Ekman, Paul}, title = {Basic Emotions}, publisher = {John Wiley and Sons}, year = {1999}, number = {3} }
Ekman, P.	Facial Expressions [BibTeX]	1999		inbook
BibTeX: @inbook{Ekman1999b, author = {Ekman, Paul}, title = {Facial Expressions}, publisher = {John Wiley and Sons}, year = {1999}, number = {16}, pages = {--} }
Ekman, P., O'Sullivan, M. & Frank, M.	A few can catch a liar [Abstract] [BibTeX]	1999	Psychological Science	article	URL
Abstract: Research suggests that most people cannot tell from demeanor when others are Eying. Such poor performance is typical not only of laypeople bur also of most professionals concerned with lying. In this study, three professional groups with special interest or skill in deception, two law-enforcement groups and a select group of clinical psychologists, obtained high accuracy in judging videotapes of people who were lying or telling the truth about their opinions. These findings strengthen earlier evidence that some professional lie catchers are highly accurate, and that behavioral clues to lying are detectable in real time. This study also provides the first evidence that some psychologists can achieve high accuracy in catching lies
BibTeX: @article{Ekman1999, author = {Ekman, Paul and O'Sullivan, M. and Frank, M.G.}, title = {A few can catch a liar}, journal = {Psychological Science}, year = {1999}, volume = {10}, number = {3}, pages = {263--266}, url = {ISI:000080735600019} }
Elbery, D.	Perspective Projection of an Ellipsoid [BibTeX]	1999		techreport	URL
BibTeX: @techreport{Elbery1999, author = {Elbery, D.}, title = {Perspective Projection of an Ellipsoid}, year = {1999}, url = {http://www.magic-software.com/Documentation PerspectiveProjectionEllipsoid.pdf} }
Elgammal, A. & Davis, L.	Probabilistic Framework for Segmenting People Under Occlusion [Abstract] [BibTeX]	2001	International Conference on Computer Vision	inproceedings	URL
Abstract: In this paper we address the problem of segmenting foregroundregions corresponding to a group of people givenmodels of their appearance that were initialized before occlusion.We present a general framework that uses maximumlikelihood estimation to estimate the best arrangementfor people in terms of 2D translation that yields a segmentationfor the foreground region. Given the segmentationresult we conduct occlusion reasoning to recover relativedepth information and we show how to utilize this depthinformation in the same segmentation framework. We alsopresent a more practical solution for the segmentation problemthat is online to avoid searching an exponential spaceof hypothesis. The person model is based on segmenting thebody into regions in order to spatially localize the color featurescorresponding to the way people are dressed. Modelingthese regions involves modeling their appearance (colordistributions) as well as their spatial distribution with respectto the body. We use a non-parametric approach basedon kernel density estimation to represent the color distributionof each region and therefore we do not restrict theclothing to be of uniform color. Instead, it can be any mixtureof colors and/or patterns. We also present a method toautomatically initialize these models and learn them beforethe occlusion.
BibTeX: @inproceedings{Elgammal2001, author = {Elgammal, Ahmed and Davis, Larry}, title = {Probabilistic Framework for Segmenting People Under Occlusion}, booktitle = {International Conference on Computer Vision}, year = {2001}, pages = {II: 145-152}, url = {http://www.cs.rutgers.edu/~elgammal/} }
Fergus, R., Perona, P. & Zisserman, A.	Object class recognition by unsupervised scale-invariant learning [BibTeX]	2003	Computer Vision and Pattern Recognition, 2003. Proceedings. 2003 IEEE Computer Society Conference on	inproceedings
BibTeX: @inproceedings{Fergus2003, author = {Fergus, R. and Perona, P. and Zisserman, A.}, title = {Object class recognition by unsupervised scale-invariant learning}, booktitle = {Computer Vision and Pattern Recognition, 2003. Proceedings. 2003 IEEE Computer Society Conference on}, year = {2003}, volume = {2}, pages = {II--264--II--271vol.2} }
Fleck, M., Forsyth, D. & Bregler, C.	Finding Naked People [Abstract] [BibTeX]	1996	European Conference on Computer Vision	conference
Abstract: This paper demonstrates an automatic system for telling whether there are naked people present in an image. The approach combines color and texture properties to obtain a mask for skin regions, which is shown to be effective for a wide range of shades and colors of skin. These skin regions are then fed to a specialized grouper, which attempts to group a human figure using geometric constraints on human structure. This approach introduces a new view of object recognition, where an object model is an organized collection of grouping hints obtained from a combination of constraints on color and texture and constraints on geometric properties such as the structure of individual parts and the relationships between parts. The system demonstrates excellent performance on a test set of 565 uncontrolled images of naked people, mostly obtained from the internet, and 4289 assorted control images, drawn from a wide collection of sources.
BibTeX: @conference{Fleck1996, author = {Fleck, Margaret and Forsyth, David and Bregler, Chris}, title = {Finding Naked People}, booktitle = {European Conference on Computer Vision}, year = {1996} }
Frigui, H.	Unsupervised learning of prototypes and attribute weights [BibTeX]	2004	Pattern Recognition	article	URL
BibTeX: @article{Frigui2004, author = {Frigui, Hichem}, title = {Unsupervised learning of prototypes and attribute weights}, journal = {Pattern Recognition}, year = {2004}, volume = {37}, pages = {567-581}, url = {http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6V14-49M0JPS-5&_user=1105409&_coverDate=03%2F31%2F2004&_rdoc=1&_fmt=&_orig=search&_sort=d&view=c&_acct=C000051666&_version=1&_urlVersion=0&_userid=1105409&md5=9e70de0f30eb6a5e52b351dc7f45383c#} }
Frigui, H. & Krishnapuram, R.	A robust competitive clustering algorithm with applications in computer vision [Abstract] [BibTeX]	1999	IEEE Transactions on Pattern Analysis and Machine Intelligence	article	DOI
Abstract: This paper addresses three major issues associated with conventional partitional clustering, namely, sensitivity to initialization, difficulty in determining the number of clusters, and sensitivity to noise and outliers. The proposed Robust Competitive Agglomeration (RCA) algorithm starts with a large number of clusters to reduce the sensitivity to initialization, and determines the actual number of clusters by a process of competitive agglomeration. Noise immunity is achieved by incorporating concepts from robust statistics into the algorithm. RCA assigns two different sets of weights for each data point: the first set of constrained weights represents degrees of sharing, and is used to create a competitive environment and to generate a fuzzy partition of the data set. The second set corresponds to robust weights, and is used to obtain robust estimates of the cluster prototypes. By choosing an appropriate distance measure in the objective function, RCA can be used to find an unknown number of clusters of various shapes in noisy data sets, as well as to fit an unknown number of parametric models simultaneously. Several examples, such as clustering/mixture decomposition, line/plane fitting, segmentation of range images, and estimation of motion parameters of multiple objects, are shown.
BibTeX: @article{Frigui1999, author = {Frigui, Hichem and Krishnapuram, Raghu}, title = {A robust competitive clustering algorithm with applications in computer vision}, journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence}, year = {1999}, volume = {21}, number = {5}, pages = {450--465}, doi = {http://dx.doi.org/10.1109/34.765656} }
Gavrila, D.	The Visual Analysis of Human Movement: A Survey [Abstract] [BibTeX]	1999	Computer Vision and Image Understanding	article	URL
Abstract: The ability to recognize humans and their activities by vision is key for a machine to interact intelligently and effortlessly with a humaninhabited environment. Because of many potentially important applications, "Looking at People" is currently one of the most active application domains in computer vision. This survey identifies a number of promising applications and provides an overview of recent developments in this domain. The scope of this survey is limited to work on whole-body or hand motion; it does not include work on human faces. The emphasis is on discussing the various methodologies; they are grouped in 2-D approaches with or without explicit shape models and 3-D approaches. Where appropriate, systems are reviewed. We conclude with some thoughts about future directions.
BibTeX: @article{Gavrila1999, author = {Gavrila, Dariu}, title = {The Visual Analysis of Human Movement: A Survey}, journal = {Computer Vision and Image Understanding}, year = {1999}, volume = {73}, number = {1}, pages = {82--98}, url = {citeseer.ist.psu.edu/gavrila99visual.html} }
Gouet, V. & Lameyre, B.	SAP: A robust approach to track objects in video streams with Snakes And Points [Abstract] [BibTeX]	2004	British Machine Vision Conference	inproceedings
Abstract: This paper presents a robust and generic approach of object tracking in videosequences. Here, the object to track is described by considering two wellknownimage primitives: first, its content is described with Points of interest.Such points are automatically extracted and then characterized according toa selective spatial appearance-based model. Second, the object envelope isdescribed with a Snake. The originality of the SAP approach consists in acomplementary use of these two primitives: the snake allows to reduce thepoints tracking to a limited area in each frame, and the spatial point descriptionis exploited during the snake tracking, making the process robust to wideocclusions. Since no model of trajectory is considered, the approach is robustto wide motions of object and camera. The relevance of this approach hasbeen evaluated on several video streams. Results obtained with the most representativeof them are presented in this paper. The algorithms involved havebeen implemented with the aim of achieving near real-time performance.
BibTeX: @inproceedings{Gouet2004, author = {Gouet, Valerie and Lameyre, Bruno}, title = {SAP: A robust approach to track objects in video streams with Snakes And Points}, booktitle = {British Machine Vision Conference}, year = {2004}, pages = {--} }
Green, P.	Trans-dimensional Markov chain Monte Carl [BibTeX]	2003	Highly Structured Stochastic Systems	incollection
BibTeX: @incollection{Green2003, author = {Green, Peter}, title = {Trans-dimensional Markov chain Monte Carl}, booktitle = {Highly Structured Stochastic Systems}, publisher = {Oxford Statistical Science Series}, year = {2003} }
Green, P.	Reversible jump Markov chain Monte Carlo computation and Bayesian model determination [Abstract] [BibTeX]	1995	Biometrika	article	URL
Abstract: This article proposes a new framework for the construction of reversible Markov chain samplers that jump between parameter subspaces of differing dimensionality, which is flexible and entirely constructive. It should therefore have wide applicability in model determination problems. The methodology is illustrated with applications to multiple change-point analysis in one and two dimensions, and to a Bayesian comparison of binomial experiments. Some key words: Change-point analysis, Image...
BibTeX: @article{Green1995, author = {Green, P.}, title = {Reversible jump Markov chain Monte Carlo computation and Bayesian model determination}, journal = {Biometrika}, year = {1995}, volume = {82}, pages = {711--732}, url = {http://citeseer.ist.psu.edu/green95reversible.html} }
Greenhill, D., Renno, J., Orwell, J. & Jones, G.	Occlusion Analysis: Learning and Utilising Depth Maps in Object Tracking [Abstract] [BibTeX]	2004	British Machine Vision Conference	inproceedings
Abstract: Complex scenes such as underground stations and malls are composed ofstatic occlusion structures such as walls, entrances, columns, turnstiles, barriers,etc. Unless this occlusion landscape is made explicit such structurescan defeat the process of tracking individuals through the scene. This paperdescribes a method of generating the probability density functions (PDFs) forthe depth of the scene at each pixel from a training set of detected blobs i.e.observations of detected moving people. As the results are necessarily noisy,a regularization process is employed to recover the most self-consistent scenedepth structure. An occlusion reasoning framework is proposed to enable objecttracking methodologies to make effective use of the recovered depth.
BibTeX: @inproceedings{Greenhill2004, author = {Greenhill, D and Renno, J and Orwell, J and Jones, G.A}, title = {Occlusion Analysis: Learning and Utilising Depth Maps in Object Tracking}, booktitle = {British Machine Vision Conference}, year = {2004}, pages = {--} }
Gruendig, M. & Hellwich, O.	3D Head Pose Estimation with Symmetry Based Illumination Model in Low Resolution Video [BibTeX]	2004	DAGM-Symposium	inproceedings
BibTeX: @inproceedings{Gruendig2004, author = {Martin Gruendig and Olaf Hellwich}, title = {3D Head Pose Estimation with Symmetry Based Illumination Model in Low Resolution Video}, booktitle = {DAGM-Symposium}, year = {2004}, pages = {45-53} }
Hamid, R., Huang, Y. & Essa, I.	ARGMode — Activity Recognition using Graphical Models [Abstract] [BibTeX]	2003	International Conference on Computer Vision and Pattern Recognition	conference
Abstract: This paper presents a new framework for tracking and recognizing complex multi-agent activities using probabilistic tracking coupled with graphical models for recognition. We employ statistical feature based particle filter to robustly track multiple objects in cluttered environments. Both color and shape characteristics are used to differentiate and track different objects so that low level visual information can be reliably extracted for recognition of complex activities. Such extracted spatio-temporal features are then used to build temporal graphical models for characterization of these activities. We demonstrate through examples in different scenarios, the generalizability and robustness of our framework.
BibTeX: @conference{Hamid2003, author = {Hamid, Raffay and Huang, Yan and Essa, Irfan}, title = {ARGMode — Activity Recognition using Graphical Models}, booktitle = {International Conference on Computer Vision and Pattern Recognition}, year = {2003} }
Hammoud, R.	Passive Eye Monitoring: Algorithms, Applications and Experiments [BibTeX]	2008		book
BibTeX: @book{Hammoud2008, author = {Hammoud, Riad}, title = {Passive Eye Monitoring: Algorithms, Applications and Experiments}, publisher = {Springer: Signals and Communication Technology}, year = {2008} }
Hammoud, R. I., Wilhelm, A., Malawey, P. & Witt, G. J.	Efficient Real-Time Algorithms for Eye State and Head Pose Tracking in Advanced Driver Support Systems [BibTeX]	2005	CVPR '05: Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) - Volume 2	inproceedings	DOI
BibTeX: @inproceedings{Hammoud2005, author = {Riad I. Hammoud and Andrew Wilhelm and Phillip Malawey and Gerald J. Witt}, title = {Efficient Real-Time Algorithms for Eye State and Head Pose Tracking in Advanced Driver Support Systems}, booktitle = {CVPR '05: Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) - Volume 2}, publisher = {IEEE Computer Society}, year = {2005}, pages = {1181}, doi = {http://dx.doi.org/10.1109/CVPR.2005.142} }
Hampel, F. R., Ronchetti, E. M., Rousseeuw, P. J. & A., S. W.	Robust Statistics: The Approach Based on Influence Functions [BibTeX]	1986		book	URL
BibTeX: @book{Hampel1986, author = {Hampel, Frank R. and Ronchetti, Elvezio M. and Rousseeuw, Peter J. and Stahel Werner A.}, title = {Robust Statistics: The Approach Based on Influence Functions}, publisher = {John Wiley & Sons Ltd}, year = {1986}, url = {http://www.wiley.com/WileyCDA/WileyTitle/productCd-0471735779.html} }
Han, B., Zhu, Y., Comaniciu, D. & Davis, L.	Kernel-Based Bayesian Filtering for Object Tracking [Abstract] [BibTeX]	2005	International Conference on Computer Vision and Pattern Recognition	inproceedings
Abstract: Particle filtering provides a general framework for propagating probability density functions in non-linear and non-Gaussian systems. However, the algorithm is based on a Monte Carlo approach and sampling is a problematic issue, especially for high dimensional problems. This paper presents a new kernelbased Bayesian filtering framework, which adopts an analytic approach to better approximate and propagate density functions. In this framework, the techniques of density interpolation and density approximation are introduced to represent the likelihood and the posterior densities by Gaussian mixtures, where all parameters such as the number of mixands, their weight, mean, and covariance are automatically determined. The proposed analytic approach is shown to perform sampling more efficiently in high dimensional space. We apply our algorithm to real-time tracking problems, and demonstrate its performance on real video sequences as well as synthetic examples.
BibTeX: @inproceedings{Han2005, author = {Han, Bohyung and Zhu, Ying and Comaniciu, Dorin and Davis, Larry}, title = {Kernel-Based Bayesian Filtering for Object Tracking}, booktitle = {International Conference on Computer Vision and Pattern Recognition}, year = {2005} }
Han, J. & Bhanu, B.	Human Activity Recognition in Thermal Infrared Imagery [BibTeX]	2005	CVPR '05: Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) - Workshops	inproceedings	DOI
BibTeX: @inproceedings{1099917, author = {Ju Han and Bir Bhanu}, title = {Human Activity Recognition in Thermal Infrared Imagery}, booktitle = {CVPR '05: Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) - Workshops}, publisher = {IEEE Computer Society}, year = {2005}, pages = {17}, doi = {http://dx.doi.org/10.1109/CVPR.2005.469} }
Haralick, R.	Computer Vision Theory: The Lack Thereof [BibTeX]	1986	Computer Vision, Graphics, and Image Processing	article
BibTeX: @article{Haralick1986, author = {Haralick, Robert}, title = {Computer Vision Theory: The Lack Thereof}, journal = {Computer Vision, Graphics, and Image Processing}, year = {1986}, number = {36}, pages = {372--386} }
Haritaoglu, I. & Flickner, M.	Detection and tracking of shopping groups in stores [Abstract] [BibTeX]	2001	International Conference on Computer Vision and Pattern Recognition	inproceedings	URL
Abstract: We describe a monocular real-time computer vision systemthat identifies shopping groups by detecting and tracking multiple people as they wait in a checkout line or servicecounter. Our system segments each frame into foregroundregions which contains multiple people. Foreground regionsare further segmented into individuals using a temporalsegmentation of foreground and motion cues. Oncea person is detected, an appearance model based on colorand edge density in conjunction with a mean-shift tracker isused to recover the person's trajectory. People are groupedtogether as a shopping group by analyzing interbody distances.The system also monitors the cashier's activities todetermine when shopping transactions start and end. Experimentalresults demonstrate the robustness and real-timeperformance of the algorithm.
BibTeX: @inproceedings{Haritaoglu2001, author = {Haritaoglu, Ismail and Flickner, Myron}, title = {Detection and tracking of shopping groups in stores}, booktitle = {International Conference on Computer Vision and Pattern Recognition}, year = {2001}, pages = {I:431-438}, url = {http://www.umiacs.umd.edu/users/hismail/} }
Haritaoglu, I. & Flickner, M.	Attentive Billboards [BibTeX]	2001	iciap	article	DOI
BibTeX: @article{Haritaoglu2001a, author = {Ismail Haritaoglu and Myron Flickner}, title = {Attentive Billboards}, journal = {iciap}, publisher = {IEEE Computer Society}, year = {2001}, volume = {00}, pages = {0162}, doi = {http://doi.ieeecomputersociety.org/10.1109/ICIAP.2001.957002} }
Haritaoglu, I., Flickner, M. & Beymer, D.	Ghost3D: Detecting Body Posture and Parts Using stereo [Abstract] [BibTeX]	2002	Workshop on Motion and Video Computing	inproceedings	URL
Abstract: This paper describes how to detect human posture and upperbody parts using overhead narrow-baseline stereo cameras.This information is extracted to understand retail customerbehavior while shopping. We propose an approachto detect body posture without using an explicit 3D humanmodel. The proposed method is based on a 3D silhouette, asilhouette-Ghost of a person, that is constructed from a 2Dsilhouette. The 2D silhouette is detected by color and disparitybackground subtraction. Once the silhouette-Ghostis generated, the head and shoulder regions are identifiedusing the topological structure of human body which constrainsthe relative position of each body part. A shape histogram,the distribution of relative positions of points ona 3D silhouette, is introduced to estimate the posture andbody parts.
BibTeX: @inproceedings{Haritaoglu2002, author = {Haritaoglu, Ismail and Flickner, Myron and Beymer, David}, title = {Ghost3D: Detecting Body Posture and Parts Using stereo}, booktitle = {Workshop on Motion and Video Computing}, year = {2002}, pages = {--}, url = {http://www.umiacs.umd.edu/users/hismail/} }
Haritaoglu, I., Harwood, D. & Davis, L.	W-4: Real-time surveillance of people and their activities [Abstract] [BibTeX]	2000	IEEE Transactions on Pattern Analysis and Machine Intelligence	article
Abstract: W-4 is a real time visual surveillance system for detecting and tracking multiple people and monitoring their activities in an outdoor environment. It operates on monocular gray-scale video imagery, or on video imagery from an infrared camera. W-4 employs a combination of shape analysis and tracking to locate people and their parts (head, hands, feet, torso) and to create models of people's appearance so that they can be tracked through interactions such as occlusions. It can determine whether a foreground region contains multiple people and can segment the region into its constituent people and track them. W-4 can also determine whether people are carrying objects, and can segment objects from their silhouettes, and construct appearance models for them so they can be identified in subsequent frames. W-4 can recognize events between people and objects, such as depositing an object, exchanging bags, or removing an object. It runs at 25 Hz for 320x240 resolution images on a 400 Mhz dual-Pentium II PC
BibTeX: @article{Haritaoglu2000, author = {Haritaoglu, Ismail and Harwood, David and Davis, Larry}, title = {W-4: Real-time surveillance of people and their activities}, journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence}, year = {2000}, volume = {22}, number = {8}, pages = {809--830} }
Hartley, R. & Zisserman, A.	Multiple View Geometry in Computer Vision [BibTeX]	2003		book
BibTeX: @book{Hartley2003, author = {Richard Hartley and Andrew Zisserman}, title = {Multiple View Geometry in Computer Vision}, publisher = {Cambridge University Press}, year = {2003} }
Havasi, L. & Sziranyi, T.	Motion Tracking Through Grouped Transient Feature Points [Abstract] [BibTeX]	2004	Advanced Concepts for Intelligent Vision Systems	inproceedings
Abstract: We introduce an object-tracking method for indefinite object shapes in case of transient motion. Moving objects of different motions sometimes overlap each other, or they partially fade into other objects in low contrast areas. We define featurepoints to be followed by optical-flow methods, and then the track of a point is described by Fourier-descriptors in a time interval.Similar motions in a neighborhood are grouped together;central feature points are defined with some interpolated motion and motion-history. The moving objects are well characterized by these central points, and multi-camera registrationis possible when motion-track components are registeredto the others.Different camera units of partly overlapping viewed area scan apply these featured trajectories for the detection and registrationof the moving objects, resulting in an automatic geometric mapping function, even in adverse lighting conditions.
BibTeX: @inproceedings{Havasi2004, author = {Havasi, Laszlo and Sziranyi, Tamas}, title = {Motion Tracking Through Grouped Transient Feature Points}, booktitle = {Advanced Concepts for Intelligent Vision Systems}, year = {2004}, pages = {--} }
Heckerman, D.	A Tutorial on Learning With Bayesian Networks [BibTeX]	1995		techreport
BibTeX: @techreport{Heckerman1995, author = {Heckerman, David}, title = {A Tutorial on Learning With Bayesian Networks}, year = {1995}, number = {MSR-TR-95-06} }
Heikkila, M., Pietikainen, M. & Heikkila, J.	A Texture-based Method for Detecting Moving Objects [Abstract] [BibTeX]	2004	British Machine Vision Conference	inproceedings
Abstract: The detection of moving objects from video frames plays an importantand often very critical role in different kinds of machine vision applicationsincluding human detection and tracking, traffic monitoring, humanmachineinterfaces and military applications, since it usually is one of thefirst phases in a system architecture. A common way to detect moving objectsis background subtraction. In background subtraction, moving objectsare detected by comparing each video frame against an existing model of thescene background. In this paper, we propose a novel block-based algorithmfor background subtraction. The algorithm is based on the Local Binary Pattern(LBP) texture measure. Each image block is modelled as a group ofweighted adaptive LBP histograms. The algorithm operates in real-time underthe assumption of a stationary camera with fixed focal length. It canadapt to inherent changes in scene background and can also handle multimodalbackgrounds.
BibTeX: @inproceedings{Heikkila2004, author = {Heikkila, M and Pietikainen, M and Heikkila, J}, title = {A Texture-based Method for Detecting Moving Objects}, booktitle = {British Machine Vision Conference}, year = {2004}, pages = {--} }
Hoaglin, D. C., Mosteller, F. & Tukey, J. W.	Understanding Robust and Exploratory Data Analysis [BibTeX]	1983		book
BibTeX: @book{Hoaglin1983, author = {Hoaglin, David C. and Mosteller, Ferderick and Tukey, John W.}, title = {Understanding Robust and Exploratory Data Analysis}, publisher = {John Wiley & Sons Ltd}, year = {1983} }
Hongeng, S.	Unsupervised Learning of Multi-Object Events [Abstract] [BibTeX]	2004	British Machine Vision Conference	inproceedings
Abstract: We present a novel approach for automatically inferring models of multiobjectevents. Objects are first detected and tracked, their motion is thensegmented into a set of primitive events. These primitive events then formthe nodes in a Markov network that encodes the entire event space. A bottomup/top-down search algorithm is developed to detect typical event structuresthat are used for classifying an observed multi-object event. We demonstrateour algorithm on clustering and inferring events in a table-laying scene.
BibTeX: @inproceedings{Hongeng2004a, author = {Hongeng, Somboon}, title = {Unsupervised Learning of Multi-Object Events}, booktitle = {British Machine Vision Conference}, year = {2004}, pages = {--} }
Hongeng, S., Nevatia, R. & Bremond, F.	Video-based event recognition: activity representation and probabilistic recognition methods [Abstract] [BibTeX]	2004	Computer Vision and Image Understanding	article
Abstract: We present a new representation and recognition method for human activities. An activity is considered to be composed of action threads, each thread being executed by a single actor. A single-thread action is represented by a stochastic finite automaton of event states, which are recognized from the characteristics of the trajectory and shape of moving blob of the actor using Bayesian methods. A multi-agent event is composed of several action threads related by temporal constraints. Multi-agent events are recognized by propagating the constraints and likelihood of event threads in a temporal logic network. We present results on real-world data and performance characterization on perturbed data.
BibTeX: @article{Hongeng2004, author = {Hongeng,Somboon and Nevatia,Ram and Bremond, Francois}, title = {Video-based event recognition: activity representation and probabilistic recognition methods}, journal = {Computer Vision and Image Understanding}, year = {2004}, volume = {96}, number = {2}, pages = {129-162} }
Horprasert, T., Harwood, D. & Davis, L.	A Statistical Approach for Real-time Robust Background Subtraction and Shadow Detection [Abstract] [BibTeX]	1999	International Conference on Computer Vision	inproceedings
Abstract: This paper presents a novel algorithm for detecting moving objects from a static backgroundscene that contains shading and shadows using color images. We develop a robust and ecientlycomputed background subtraction algorithm that is able to cope with local illumination changes,such as shadows and highlights, as well as global illumination changes. The algorithm is basedon a proposed computational color model which separates the brightness from the chromaticitycomponent. We have applied this method to real image sequences of both indoor and outdoorscenes. The results, which demonstrate the system's performance, and some speed up techniqueswe employed in our implementation are also shown.
BibTeX: @inproceedings{Horprasert1999, author = {Horprasert, Thanarat and Harwood, David and Davis, Larry}, title = {A Statistical Approach for Real-time Robust Background Subtraction and Shadow Detection}, booktitle = {International Conference on Computer Vision}, year = {1999}, pages = {1-19} }
Hosie, R., Venkatesh, S. & West, G.	Classifying and Detecting Group Behaviour from Visual Surveillance Data [BibTeX]	1998	ICPR '98: Proceedings of the 14th International Conference on Pattern Recognition-Volume 1	inproceedings
BibTeX: @inproceedings{Hosie1998, author = {Robin Hosie and Svetha Venkatesh and Geoff West}, title = {Classifying and Detecting Group Behaviour from Visual Surveillance Data}, booktitle = {ICPR '98: Proceedings of the 14th International Conference on Pattern Recognition-Volume 1}, publisher = {IEEE Computer Society}, year = {1998}, pages = {602} }
Hui, S. K., Fader, P. & Bradlow, E.	Path Data in Marketing: An Integrative Framework and Prospectus for Model-Building [BibTeX]	2007	Social Science Research Network	article
BibTeX: @article{Hui2007, author = {Hui, Sam K. and Fader, Peter and Bradlow, Eric}, title = {Path Data in Marketing: An Integrative Framework and Prospectus for Model-Building}, journal = {Social Science Research Network}, year = {2007} }
Isard, M. & Blake, A.	Condensation -- conditional density propagation for visual tracking [BibTeX]	1998	International Journal of Computer Vision	article	URL
BibTeX: @article{Isard1998, author = {Isard, Michael and Blake, Andrew}, title = {Condensation -- conditional density propagation for visual tracking}, journal = {International Journal of Computer Vision}, year = {1998}, volume = {29}, pages = {5-28}, url = {citeseer.ist.psu.edu/isard98condensation.html} }
Isard, M. & MacCormick, J.	BraMBLe: A Bayesian Multiple-Blob Tracker [Abstract] [BibTeX]	2001	International Conference on Computer Vision	inproceedings
Abstract: Blob trackers have become increasingly powerful in recentyears largely due to the adoption of statistical appearancemodels which allow effective background subtraction androbust tracking of deforming foreground objects. It has beenstandard, however, to treat background and foregroundmodelling as separate processes - background subtractionis followed by blob detection and tracking-which preventsa principled computation of image likelihoods. This paperpresents two theoretical advances which address this limitationand lead to a robust multiple-person tracking systemsuitable for single-camera real-time surveillance applications.The first innovation is a multi-blob likelihood functionwhich assigns directly comparable likelihoods to hypothesescontaining different numbers of objects. This likelihoodfunction has a rigorous mathematical basis: it is adaptedfrom the theory of Bayesian correlation, but uses the assumptionof a static camera to create a more specific backgroundmodel while retaining a unified approach to backgroundand foreground modelling. Second we introduce aBayesian filter for tracking multiple objects when the numberof objects present is unknown and varies over time. Weshow how a particle filter can be used to perform joint inferenceon both the number of objects present and their con-figurations. Finally we demonstrate that our system runscomfortably in real time on a modest workstation when thenumber of blobs in the scene is small.
BibTeX: @inproceedings{Isard2001, author = {Isard, Michael and MacCormick, John}, title = {BraMBLe: A Bayesian Multiple-Blob Tracker}, booktitle = {International Conference on Computer Vision}, year = {2001}, volume = {2}, pages = {34-41} }
Itti, L.	Models of Bottom-Up Attention and Saliency [Abstract] [BibTeX]	2005	Neurobiology of Attention	incollection
Abstract: Visually conspicuous, or so-called salient, stimuli often have the capability of attracting focal visual attention towards their locations. Several computational architectures subserving this bottom-up, stimulus-driven, spatiotemporal deployment of attention are reviewed in this article. The resulting computational models have applications not only to the prediction of visual search psychophysics, but also, in the domain of machine vision, to the rapid selection of regions of interest in complex, cluttered visual environments. We describe an unusal such application, to the objective evaluation of advertising designs.
BibTeX: @incollection{Itti2005, author = {L. Itti}, title = {Models of Bottom-Up Attention and Saliency}, booktitle = {Neurobiology of Attention}, publisher = {Elsevier}, year = {2005}, pages = {576-582} }
Jain, A. & Dubes, R.	Algorithms for clustering data [BibTeX]	1988		book
BibTeX: @book{Jain1988, author = {Jain, Anil and Dubes, Richard}, title = {Algorithms for clustering data}, publisher = {Prentice-Hall}, year = {1988} }
Jedynak, B., Zheng, H. & Daoudi, M.	Statistical Models for Skin Detection [Abstract] [BibTeX]	2003	Workshop on Statistical Analysis in Computer Vision	conference
Abstract: We consider a sequence of three models for skin detection built from a large collection of labelled images. Each model is a maximum entropy model with respect to constraints concerning marginal distributions. Our models are nested. The first model is well known from practitioners. Pixels are considered as independent. The second model is a Hidden Markov Model. It includes constraints that force smoothness of the solution. The third model is a first order model. The full color gradient is included. Parameter estimation as well as optimization cannot be tackled without approximations. We use thoroughly Bethe tree approximation of the pixel lattice. Within it , parameter estimation is eradicated and the belief propagation algorithm permits to obtain exact and fast solution for skin probability at pixel locations. We then assess the performance on the Compaq database.
BibTeX: @conference{Jedynak2003, author = {Jedynak, Bruno and Zheng, Huicheng and Daoudi, Mohamed}, title = {Statistical Models for Skin Detection}, booktitle = {Workshop on Statistical Analysis in Computer Vision}, year = {2003} }
Johnson, S.	Emergence: The Connected Lives of Ants, Brains, Cities, and Software [BibTeX]	2001		book
BibTeX: @book{Johnson2001, author = {Johnson, Steven}, title = {Emergence: The Connected Lives of Ants, Brains, Cities, and Software}, publisher = {Scribner}, year = {2001} }
Jones, M. J. & Rehg, J. M.	Statistical Color Models with Application to Skin Detection [Abstract] [BibTeX]	2002	International Journal of Computer Vision	article	URL
Abstract: The existence of large image datasets such as the set of photos on the World Wide Web make it possible to build powerful generic models for low-level image attributes like color using simple histogram learning techniques. We describe the construction of color models for skin and non-skin classes from a dataset of nearly 1 billion labelled pixels. These classes exhibit a surprising degree of separability which we exploit by building a skin pixel detector achieving a detection rate of 80% with 8.5% false positives. We compare the performance of histogram and mixture models in skin detection and find histogram models to be superior in accuracy and computational cost. Using aggregate features computed from the skin pixel detector we build a surprisingly effective detector for naked people. Our results suggest that color can be a more powerful cue for detecting people in unconstrained imagery than was previously suspected. We believe this work is the most comprehensive and detailed exploration of skin color models to date.
BibTeX: @article{Jones2002, author = {Jones, Michael J. and Rehg, James M.}, title = {Statistical Color Models with Application to Skin Detection}, journal = {International Journal of Computer Vision}, year = {2002}, volume = {46}, number = {1}, pages = {81-96}, url = {http://citeseer.ist.psu.edu/jones99statistical.html} }
Kahl, F. & Heyden, A.	Affine Structure and Motion from Points, Lines and Conics [BibTeX]	1999	International Journal of Computer Vision	article	URL
BibTeX: @article{Kahl1999, author = {Fredrik Kahl and Anders Heyden}, title = {Affine Structure and Motion from Points, Lines and Conics}, journal = {International Journal of Computer Vision}, publisher = {Springer Netherlands}, year = {1999}, volume = {33}, number = {3}, pages = {163--180}, url = {citeseer.ist.psu.edu/365175.html} }
Kam, A., Ann, T., Lung, E., Yun, Y. & Junxian, W.	Automated Recognition of Highly Complex Human Behavior [Abstract] [BibTeX]	2004	International Conference on Pattern Recognition	conference
Abstract: We describe a framework for automated complex human behavior recognition, illustrating important concepts with specific examples drawn from our work on a unique platform designed to understand water crises related behaviors in a public swimming pool. We argue for a hierarchical representation, leveraging on quantitative descriptors to model a behaviour’s intermediate semantics. Complex behaviour inference is then demonstrated using a novel regression-based approach based on a modified version of a functional link network which learns quickly and classifies accurately in comparison with other competing decision making schemes.
BibTeX: @conference{Kam2004, author = {Kam, Alvin and Ann, Toh and Lung, Eng and Yun, Yau and Junxian, Wang}, title = {Automated Recognition of Highly Complex Human Behavior}, booktitle = {International Conference on Pattern Recognition}, year = {2004} }
Kemp, C. & Drummond, T.	Multi-Modal Tracking using Texture Changes [Abstract] [BibTeX]	2004	British Machine Vision Conference	inproceedings
Abstract: We present a method for efficiently generating a representation of a multimodalposterior probability distribution. The technique combines ideas fromRANSAC and particle filtering such that the 3D visual tracking problem can bepartitioned into two levels, while maintaining multiple hypotheses throughout.A simple texture change-point detector finds multiple hypotheses for theposition of image edgels. From these, multiple locations for each scene edgeare generated. Finally we determine the best pose of the whole structure.While the multi-modal representation is strongly related to particle filteringtechniques, this approach is driven by data from the image. Hence the resultingsystem is able to perform robust visual tracking of all six degrees offreedom in real time. Real video sequences are used to compare the completetracking system to previous systems.
BibTeX: @inproceedings{Kemp2004, author = {Kemp, Christopher and Drummond, Tom}, title = {Multi-Modal Tracking using Texture Changes}, booktitle = {British Machine Vision Conference}, year = {2004}, pages = {--} }
Khan, S. & Shah, M.	Consistent Labeling of Tracked Objects in Multiple Cameras with Overlapping Fields of View [BibTeX]	2003	IEEE Transactions on Pattern Analysis and Machine Intelligence	article	DOI
BibTeX: @article{Khan2003, author = {Sohaib Khan and Mubarak Shah}, title = {Consistent Labeling of Tracked Objects in Multiple Cameras with Overlapping Fields of View}, journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence}, publisher = {IEEE Computer Society}, year = {2003}, volume = {25}, number = {10}, pages = {1355-1360}, doi = {http://doi.ieeecomputersociety.org/10.1109/TPAMI.2003.1233912} }
Khan, Z.	MCMC-Based Particle Filtering for Tracking a Variable Number of Interacting Targets [BibTeX]	2005	IEEE Trans. Pattern Anal. Mach. Intell.	article	DOI
BibTeX: @article{Khan2005a, author = {Zia Khan}, title = {MCMC-Based Particle Filtering for Tracking a Variable Number of Interacting Targets}, journal = {IEEE Trans. Pattern Anal. Mach. Intell.}, publisher = {IEEE Computer Society}, year = {2005}, volume = {27}, number = {11}, pages = {1805--1918}, note = {Member-Tucker Balch and Member-Frank Dellaert}, doi = {http://dx.doi.org/10.1109/TPAMI.2005.223} }
Khan, Z., Balch, T. & Dellaert, F.	Multitarget Tracking with Split and Merged Measurements [Abstract] [BibTeX]	2005	International Conference on Computer Vision and Pattern Recognition	conference	URL
Abstract: In many multitarget tracking applications in computer vision, a detection algorithm provides locations of potential targets. Subsequently, the measurements are associated with previously estimated target trajectories in a data association step. The output of the detector is often imperfect and the detection data may include multiple, split measurements from a single target or a single merged measurement from several targets. To address this problem, we introduce a multiple hypothesis tracker for interacting targets that generate split and merged measurements. The tracker is based on an efficient Markov chain Monte Carlo (MCMC) based auxiliary variable particle filter. The particle filter is Rao-Blackwellized such that the continuous target state parameters are estimated analytically, and an MCMC sampler generates samples from the large discrete space of data associations. In addition, we include experimental results in a scenario where we track several interacting targets that generate these split and merged measurements.
BibTeX: @conference{Khan2005, author = {Khan, Zia and Balch,Tucker and Dellaert, Frank}, title = {Multitarget Tracking with Split and Merged Measurements}, booktitle = {International Conference on Computer Vision and Pattern Recognition}, year = {2005}, url = {http://www.cc.gatech.edu/~dellaert/} }
Kim, G., Huber, D. & Hebert, M.	Segmentation of Salient Regions in Outdoor Scenes Using Imagery and 3-D Data [BibTeX]	2008	IEEE Workshop on Applications of Computer VIsion (WACV08)	inproceedings
BibTeX: @inproceedings{Kim2008, author = {Gunhee Kim and Daniel Huber and Martial Hebert}, title = {Segmentation of Salient Regions in Outdoor Scenes Using Imagery and 3-D Data}, booktitle = {IEEE Workshop on Applications of Computer VIsion (WACV08)}, publisher = {IEEE Computer Society}, year = {2008} }
Kim, K., Chalidabhongse, T., Harwood, D. & Davis, L.	Background Modeling and Subtraction by Codebook Construction [Abstract] [BibTeX]	2004	International Conference on Image Processing	inproceedings	URL
Abstract: We present a new fast algorithm for background modelingand subtraction. Sample background values at each pixelare quantized into codebooks which represent a compressedform of background model for a long image sequence. Thisallows us to capture structural background variation due toperiodic-like motion over a long period of time under limitedmemory. Our method can handle scenes containingmoving backgrounds or illumination variations (shadows andhighlights), and it achieves robust detection for compressedvideos. We compared our method with other multimodemodeling techniques.
BibTeX: @inproceedings{Kim2004, author = {Kim, Kyungnam and Chalidabhongse, Thanarat and Harwood, David and Davis, Larry}, title = {Background Modeling and Subtraction by Codebook Construction}, booktitle = {International Conference on Image Processing}, year = {2004}, volume = {5}, pages = {3061- 3064}, url = {http://www.umiacs.umd.edu/~knkim/UMD-BGS/} }
Klawonn, F. & Höppner, F.	What Is Fuzzy about Fuzzy Clustering? Understanding and Improving the Concept of the Fuzzifier. [BibTeX]	2003	Advances in Intelligent Data Analysis V	inproceedings	URL
BibTeX: @inproceedings{Klawonn2003, author = {Klawonn, Frank and Höppner, Frank}, title = {What Is Fuzzy about Fuzzy Clustering? Understanding and Improving the Concept of the Fuzzifier.}, booktitle = {Advances in Intelligent Data Analysis V}, year = {2003}, volume = {2810}, pages = {254-264}, url = {http://www.springerlink.com/content/wywbmnub0p4dbleb/} }
Kushal, A., Bansal, V. & Banerjee, S.	A simple method for interactive 3D reconstruction and camera calibration from a single view [Abstract] [BibTeX]	2005	Indian Conference on Computer Vision, Graphics and Image Processing	inproceedings
Abstract: We present a simple and intuitive method for interactive 3Dreconstruction and camera calibration from a single imageof a structured scene. The method is based on manual registrationof two world planes. We present experimental resultson some test images.
BibTeX: @inproceedings{Kushal2005, author = {Kushal, Akash and Bansal, Vikas and Banerjee, Subhashis}, title = {A simple method for interactive 3D reconstruction and camera calibration from a single view}, booktitle = {Indian Conference on Computer Vision, Graphics and Image Processing}, year = {2005}, pages = {--} }
LaViola, J.	A Comparison of Unscented and Extended Kalman Filtering for Estimating Quaternion Motion [Abstract] [BibTeX]	2003	American Control Conference	conference
Abstract: The unscented Kalman filter is a superior alternative to the extended Kalman filter for a variety of estimation and control problems. However, its effectiveness for improving human motion tracking for virtual reality applications in the presence of noisy data has been unexplored. In this paper, we present an empirical study comparing the performance of unscented and extended Kalman filtering for improving human head and hand tracking. Specifically, we examine human head and hand orientation motion signals, represented with quaternions, which are critical for correct viewing perspectives in virtual reality. Our experimental results and analysis indicate that unscented Kalman filtering performs equivalently with extended Kalman filtering. However, the additional computational overhead of the unscented Kalman filter and quasi-linear nature of the quaternion dynamics lead to the conclusion that the extended Kalman filter is a better choice for estimating quaternion motion in virtual reality applications.
BibTeX: @conference{LaViola2003, author = {LaViola, Joseph}, title = {A Comparison of Unscented and Extended Kalman Filtering for Estimating Quaternion Motion}, booktitle = {American Control Conference}, year = {2003} }
Lee, M. W., Cohen, I. & Jung, S. K.	Particle Filter with Analytical Inference for Human Body Tracking [Abstract] [BibTeX]	2002	Workshop on Motion and Video Computing	inproceedings
Abstract: This paper introduces a framework that integrates analytical inference into the particle filtering scheme for human body tracking. The analytical inference is provided by body parts detection, and is used to update subsets of state parameters representing the human pose. This reduces the degree of randomness and decreases the required number of particles. This new technique is a significant improvement over the standard particle filtering with the advantages of performing automatic track initialization, recovering from tracking failures, and reducing the computational load.
BibTeX: @inproceedings{Lee2002, author = {Lee, Mun Wai and Cohen, Isaac and Jung, Soon Ki}, title = {Particle Filter with Analytical Inference for Human Body Tracking}, booktitle = {Workshop on Motion and Video Computing}, year = {2002}, pages = {--} }
Leykin, A.	Visual Human Tracking and Group Activity Analysis: A Video Mining System for Retail Marketing [BibTeX]	2007	School: Indiana University	phdthesis
BibTeX: @phdthesis{Leykin2007a, author = {Leykin, Alex}, title = {Visual Human Tracking and Group Activity Analysis: A Video Mining System for Retail Marketing}, school = {Indiana University}, year = {2007} }
Leykin, A. & Hammoud, R.	Robust Multi-Pedestrian Tracking in Thermal-Visible Surveillance Videos [BibTeX]	2006	cvprw	inproceedings	DOI
BibTeX: @inproceedings{Leykin2006, author = {Alex Leykin and Riad Hammoud}, title = {Robust Multi-Pedestrian Tracking in Thermal-Visible Surveillance Videos}, journal = {cvprw}, publisher = {IEEE Computer Society}, year = {2006}, volume = {0}, pages = {136}, doi = {http://doi.ieeecomputersociety.org/10.1109/CVPRW.2006.175} }
Leykin, A., Ran, Y. & Hammoud, R.	Thermal-Visible Video Fusion for Moving Target Tracking and Pedestrian Classification [BibTeX]	2007	Object Tracking and Classification in and Beyond the Visible Spectrum	inproceedings
BibTeX: @inproceedings{Leykin2007, author = {Leykin, Alex and Ran, Yang and Hammoud, Riad}, title = {Thermal-Visible Video Fusion for Moving Target Tracking and Pedestrian Classification}, booktitle = {Object Tracking and Classification in and Beyond the Visible Spectrum}, year = {2007}, pages = {1-8} }
Leykin, A. & Tuceryan, M.	Detecting shopper groups in video sequences [BibTeX]	2007	AVSS	inproceedings
BibTeX: @inproceedings{Leykin2007b, author = {Leykin, A. and Tuceryan, M.}, title = {Detecting shopper groups in video sequences}, booktitle = {AVSS}, year = {2007}, pages = {417-422} }
Lien, J., Kanade, T., Cohn, J. & Li, C.	Detection, Tracking, and Classification of Action Units in Facial Expression [Abstract] [BibTeX]	1999	Journal of Robotics and Autonomous Systems	article	URL
Abstract: Most of the current work on automated facial expression analysis attempt to recognize asmall set of prototypic expressions, such as joy and fear. Such prototypic expressions, however,occur infrequently, and human emotions and intentions are communicated more often bychanges in one or two discrete features. To capture the full range of facial expression, detection,tracking, and classification of fine-grained changes in facial features are needed. We developedthe first version of a computer vision system that is sensitive to subtle changes in the face. Thesystem includes three modules to extract feature information: dense-flow extraction using awavelet motion model, facial feature tracking, and edge and line extraction. The featureinformation thus extracted is fed to discriminant classifiers or hidden Markov models thatclassify it into FACS action units, the descriptive system to code fine-grained changes in facialexpression. The system was tested on image sequences from 100 male and female subjects ofvaried ethnicity. Agreement with manual FACS coding was strong for the results based ondense-flow extraction and facial feature tracking, and strong to moderate for edge and lineextraction.
BibTeX: @article{Lien1999, author = {Lien, James and Kanade, Takeo and Cohn, Jeffrey and Li, Ching-Chung}, title = {Detection, Tracking, and Classification of Action Units in Facial Expression}, journal = {Journal of Robotics and Autonomous Systems}, year = {1999}, pages = {--}, url = {http://www-2.cs.cmu.edu/~face/Papers/RA_Journal99b.PDF} }
Liu, Z. & Sarkar, S.	Challenges in Segmentation of Human Forms in Outdoor Video [Abstract] [BibTeX]	2004	International Conference on Computer Vision and Pattern Recognition Workshop	inproceedings
Abstract: Most gait and activity recognition algorithms rely on the use of silhouettes as the low-level representation. However, the detection of good silhouettes is still an open problem, particularly for sequences that are taken outdoors. Illumination conditions, compression artifacts, and low number of pixels on the subject, contribute to the difficulty. Presently, these issues are either ignored by using indoor data or addressed, on a case by case basis, by employing, essentially, a "bag of tricks" based approach. We argue for a more formal approach, based on generic shape and motion models to handle a variety of these problems, under the umbrella of one formalism. We present an HMM-based Eigen Stance model, built based on manually created silhouettes from 71 individuals. The population HMM helps map a frame in any given sequence to a stance and the appearance based Eigen-Stance model is used to reconstruct the computed silhouette in that frame. We quantify the performance in terms of signal based criteria of missed detection and false positive prediction rate. We also show results on three different databases.
BibTeX: @inproceedings{Liu2004, author = {Liu, Zongyi and Sarkar, Sudeep}, title = {Challenges in Segmentation of Human Forms in Outdoor Video}, booktitle = {International Conference on Computer Vision and Pattern Recognition Workshop}, year = {2004} }
Low, K. & Ilie, A.	Computing a View Frustum to Maximize an Object's Image Area [Abstract] [BibTeX]	2003	#j-J-GRAPHICS-TOOLS#	article	URL
Abstract: This paper presents a method to compute a view frustum for a 3D object viewed from a given viewpoint, such that the object is completely enclosed in the frustum and the object's image area is also near-maximal in the given 2D rectangular viewing region. This optimization can be used to improve the resolution of shadow maps and texture maps for projective texture mapping. Instead of doing the optimization in 3D space to find a good view frustum, our method uses a 2D approach. The basic idea of our approach is as follows. First, from the given viewpoint, a conveniently-computed view frustum is used to project the 3D vertices of the object to their corresponding 2D image points. A tight 2D bounding quadrilateral is then computed to enclose these 2D image points. Next, considering the projective warp between the bounding quadrilateral and the rectangular viewing region, our method applies a technique of camera calibration to compute a new view frustum that generates an image that covers the viewing region as much as possible.
BibTeX: @article{Low2003, author = {Kok-Lim Low and Adrian Ilie}, title = {Computing a View Frustum to Maximize an Object's Image Area}, journal = {#j-J-GRAPHICS-TOOLS#}, year = {2003}, volume = {8}, number = {1}, pages = {3--15}, url = {http://www.acm.org/jgt/papers/LowIlie03/} }
Lowe, D. G.	Distinctive Image Features from Scale-Invariant Keypoints [BibTeX]	2004	Int. J. Comput. Vision	article	DOI
BibTeX: @article{Lowe2004, author = {David G. Lowe}, title = {Distinctive Image Features from Scale-Invariant Keypoints}, journal = {Int. J. Comput. Vision}, publisher = {Kluwer Academic Publishers}, year = {2004}, volume = {60}, number = {2}, pages = {91--110}, doi = {http://dx.doi.org/10.1023/B:VISI.0000029664.99615.94} }
Micolotta, A. & Bowden, R.	View-based Location and Tracking of Body Parts for Visual Interaction [Abstract] [BibTeX]	2004	British Machine Vision Conference	inproceedings
Abstract: This paper presents a real time approach to locate and track the uppertorso of the human body. Our main interest is not in 3D biometric accuracy,but rather a sufficient discriminatory representation for visual interaction.The algorithm employs background suppression and a general approximationto body shape, applied within a particle filter framework, making use ofintegral images to maintain real-time performance. Furthermore, we presenta novel method to disambiguate the hands of the subject and to predict thelikely position of elbows. The final system is demonstrated segmenting multiplesubjects from a cluttered scene at above real time operation.
BibTeX: @inproceedings{Micolotta2004, author = {Micolotta, Antonio and Bowden, Richard}, title = {View-based Location and Tracking of Body Parts for Visual Interaction}, booktitle = {British Machine Vision Conference}, year = {2004}, pages = {--} }
Mittal, A. & Davis, L.	M2Tracker: A Multi-view Approach to Segmenting and Tracking People in a Cluttered Scene Using Region-Based Stereo [BibTeX]	2002	European Conference on Computer Vision	inproceedings
BibTeX: @inproceedings{Mittal2002, author = {Mittal, Anurag and Davis, Larry}, title = {M2Tracker: A Multi-view Approach to Segmenting and Tracking People in a Cluttered Scene Using Region-Based Stereo}, booktitle = {European Conference on Computer Vision}, year = {2002}, volume = {51}, number = {3}, pages = {189--203} }
Mo?nne-Loccoz, N., Br?mond, F. & Thonnat, M.	Recurrent Bayesian Network for the Recognition of Human Behaviors from Video [Abstract] [BibTeX]	2003	International Conference on Computer Vision Systems	conference
Abstract: We propose an original bayesian approach to recognize human behaviors from video streams. Mobile objects and their visual features are computed by a vision module. Then, using a Recurrent Bayesian Network, behaviors of the mobile objects are recognized through the temporal evolution of their visual features.
BibTeX: @conference{Mo?nne-Loccoz2003, author = {Mo?nne-Loccoz, Nicolas and Br?mond, Fran?ois and Thonnat, Monique}, title = {Recurrent Bayesian Network for the Recognition of Human Behaviors from Video}, booktitle = {International Conference on Computer Vision Systems}, year = {2003} }
Moeslund, T. B., Hilton, A. & Kr&252;ger, V.	A survey of advances in vision-based human motion capture and analysis [BibTeX]	2006	Computer Vision and Image Understanding	article	DOI URL
BibTeX: @article{Moeslund2006, author = {Moeslund, Thomas B. and Hilton, Adrian and Kr&252;ger, Volker}, title = {A survey of advances in vision-based human motion capture and analysis}, journal = {Computer Vision and Image Understanding}, publisher = {Elsevier Science Inc.}, year = {2006}, volume = {104}, number = {2}, pages = {90--126}, url = {http://dx.doi.org/10.1016/j.cviu.2006.08.002}, doi = {http://dx.doi.org/10.1016/j.cviu.2006.08.002} }
Morariu, V. I. & Camps, O. I.	Modeling Correspondences for Multi-Camera Tracking Using Nonlinear Manifold Learning and Target Dynamics [BibTeX]	2006	cvpr	inproceedings	DOI
BibTeX: @inproceedings{Morariu2006, author = {Vlad I. Morariu and Octavia I. Camps}, title = {Modeling Correspondences for Multi-Camera Tracking Using Nonlinear Manifold Learning and Target Dynamics}, journal = {cvpr}, publisher = {IEEE Computer Society}, year = {2006}, volume = {1}, pages = {545-552}, doi = {http://doi.ieeecomputersociety.org/10.1109/CVPR.2006.189} }
Morris, D. & Rehg, J.	Singularity Analysis for Articulated Object Tracking [Abstract] [BibTeX]	1998	International Conference on Computer Vision and Pattern Recognition	conference
Abstract: We analyze the use of kinematic constraints for articulated object tracking. Conditions for the occurrence of singularities in 3-D models are presented and their effects on tracking are characterized. We describe a novel 2-D Scaled Prismatic Model (SPM) for figure registration. In contrast to 3-D kinematic models, the SPM has fewer singularity problems and does not require detailed knowledge of the 3-D kinematics. We fully characterize the singularities in the SPM and illustrate tracking through singularities using synthetic and real examples with 3-D and 2-D models. Our results demonstrate the significant benefits of the SPM in tracking with a single source of video.
BibTeX: @conference{Morris1998, author = {Morris, Daniel and Rehg, James}, title = {Singularity Analysis for Articulated Object Tracking}, booktitle = {International Conference on Computer Vision and Pattern Recognition}, publisher = {IEEE Computer Society}, year = {1998}, pages = {289} }
Navalpakkam, V. & Itti, L.	Search goal tunes visual features optimally. [BibTeX]	2007 Feb 15	Neuron	article	DOI
BibTeX: @article{Navalpakkam2007Feb15, author = {Navalpakkam, V and Itti, L}, title = {Search goal tunes visual features optimally.}, journal = {Neuron}, year = {2007 Feb 15}, volume = {53}, number = {4}, pages = {605-17}, doi = {http://dx.doi.org/10.1016/j.neuron.2007.01.018} }
Navalpakkam, V. & Itti, L.	An Integrated Model of Top-Down and Bottom-Up Attention for Optimizing Detection Speed [BibTeX]	2006	CVPR '06: Proceedings of the 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition	inproceedings	DOI
BibTeX: @inproceedings{Navalpakkam2006, author = {Vidhya Navalpakkam and Laurent Itti}, title = {An Integrated Model of Top-Down and Bottom-Up Attention for Optimizing Detection Speed}, booktitle = {CVPR '06: Proceedings of the 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition}, publisher = {IEEE Computer Society}, year = {2006}, pages = {2049--2056}, doi = {http://dx.doi.org/10.1109/CVPR.2006.54} }
Navalpakkam, V. & Itti, L.	Modeling the influence of task on attention [Abstract] [BibTeX]	2005	Vision Research	article	DOI
Abstract: We propose a computational model for the task-specific guidance of visual attention in real-world scenes. Our model emphasizes four aspects that are important in biological vision: determining task-relevance of an entity, biasing attention for the low-level visual features of desired targets, recognizing these targets using the same low-level features, and incrementally building a visual map of task-relevance at every scene location. Given a task definition in the form of keywords, the model first determines and stores the task-relevant entities in working memory, using prior knowledge stored in long-term memory. It attempts to detect the most relevant entity by biasing its visual attention system with the entity's learned low-level features. It attends to the most salient location in the scene, and attempts to recognize the attended object through hierarchical matching against object representations stored in long-term memory. It updates its working memory with the task-relevance of the recognized entity and updates a topographic task-relevance map with the location and relevance of the recognized entity. The model is tested on three types of tasks: single-target detection in 343 natural and synthetic images, where biasing for the target accelerates target detection over two-fold on average; sequential multiple-target detection in 28 natural images, where biasing, recognition, working memory and long term memory contribute to rapidly finding all targets; and learning a map of likely locations of cars from a video clip filmed while driving on a highway. The model's performance on search for single features and feature conjunctions is consistent with existing pyschophysical data. These results of our biologically-motivated architecture suggest that the model may provide a reasonable approximation to many brain processes involved in complex task-driven visual behaviors.
BibTeX: @article{Navalpakkam2005, author = {V. Navalpakkam and L. Itti}, title = {Modeling the influence of task on attention}, journal = {Vision Research}, year = {2005}, volume = {45}, number = {2}, pages = {205-231}, doi = {http://dx.doi.org/10.1016/j.visres.2004.07.042} }
Okuma, K., Taleghani, A., De Freitas, N., Little, J. & Lowe, D.	A boosted particle filter: Multitarget detection and tracking [Abstract] [BibTeX]	2004	European Conference on Computer Vision	conference	URL
Abstract: The problem of tracking a varying number of non-rigid objects has two major dificulties. First, the observation models and target distributions can be highly non-linear and non-Gaussian. Second, the presence of a large, varying number of objects creates complex interactions with overlap and ambiguities. To surmount these diculties, we introduce a vision system that is capable of learning, detecting and tracking the objects of interest. The system is demonstrated in the context of tracking hockey players using video sequences. Our approach combines the strengths of two successful algorithms: mixture particle lters and Adaboost. The mixture particle lter [17] is ideally suited to multi-target tracking as it assigns a mixture component to each player. The crucial design issues in mixture particle lters are the choice of the proposal distribution and the treatment of objects leaving and entering the scene. Here, we construct the proposal distribution using a mixture model that incorporates information from the dynamic models of each player and the detection hypotheses generated by Adaboost. The learned Adaboost proposal distribution allows us to quickly detect players entering the scene, while the ltering process enables us to keep track of the individual players. The result of interleaving Adaboost with mixture particle lters is a simple, yet powerful and fully automatic multiple object tracking system.
BibTeX: @conference{Okuma2004, author = {Okuma, Kenji and Taleghani, Ali and De Freitas,Nando and Little, James and Lowe, David}, title = {A boosted particle filter: Multitarget detection and tracking}, booktitle = {European Conference on Computer Vision}, year = {2004}, url = {http://www.cs.ubc.ca/~lowe} }
Oliver, N.	Towards Perceptual Intelligence: Statistical Modeling of Human Individual and Interactive Behaviors [Abstract] [BibTeX]	2000	School: Massachusetts Institute of Technology	phdthesis
Abstract: This thesis presents a computational framework for the automatic recognition and prediction of dierent kinds of human behaviors from video cameras and other sensors, via perceptually intelligent systems that automatically sense and correctly classify human behaviors, by means of Machine Per- ception and Machine Learning techniques. In the thesis I develop the statistical machine learning algorithms (dynamic graphical models) necessary for detecting and recognizing individual and interactive behaviors. In the case of the interactions two Hidden Markov Models (HMMs) are coupled in a novel architecture called Coupled Hidden Markov Models (CHMMs) that explicitly captures the interactions between them. The algorithms for learning the parameters from data as well as for doing inference with those models are developed and described. Four systems that experimentally evaluate the proposed paradigm are presented: (1) LAFTER, an automatic face detection and tracking system with facial expression recognition; (2) a Tai-Chi gesture recognition system; (3) a pedestrian surveillance system that recognizes typical human to human interactions; (4) and a SmartCar for driver maneuver recognition. These systems capture human behaviors of dierent nature and increasing complexity: rst, isolated, single-user facial expressions, then, two-hand gestures and human-to-human interactions, and nally complex behaviors where human performance is mediated by a machine, more specically, a car. The metric that is used for quantifying the quality of the behavior models is their accuracy: howwell they are able to recognize the behaviors on testing data. Statisticalmachine learning usually suers from lack of data for estimating all the parameters in the models. In order to alleviate this problem, synthetically generated data are used to bootstrap the models creating 'prior models' that are further trained using much less real data than otherwise it would be required. The Bayesian nature of the approach let us do so. The predictive power of these models lets us categorize human actions very soon after the beginning of the action. Because of the generic nature of the typical behaviors of each of the implemented systems there is a reason to believe that this approach to modeling human behavior would generalize to other dynamic human-machine systems. This would allow us to recognize automatically people's intended action, and thus build control systems that dynamically adapt to suit the human's purposes better.
BibTeX: @phdthesis{Oliver2000a, author = {Oliver, Nuria}, title = {Towards Perceptual Intelligence: Statistical Modeling of Human Individual and Interactive Behaviors}, school = {Massachusetts Institute of Technology}, year = {2000} }
Oliver, N., Horvitz, E. & Garg, A.	Layered Representations for Human Activity Recognition [Abstract] [BibTeX]	2002	International Conference on Multimodal Interfaces	conference	URL
Abstract: We present the use of layered probabilistic representations using Hidden Markov Models for performing sensing, learning, and inference at multiple levels of temporal granularity. We describe the use of the representation in a system that diagnoses states of a user’s activity based on real-time streams of evidence from video, acoustic, and computer interactions. We review the representation, present an implementation, and report on experiments with the layered representation in an office-awareness application.
BibTeX: @conference{Oliver2002, author = {Oliver, N. and Horvitz, E. and Garg, A.}, title = {Layered Representations for Human Activity Recognition}, booktitle = {International Conference on Multimodal Interfaces}, year = {2002}, pages = {3--8}, url = {citeseer.ist.psu.edu/oliver02layered.html} }
Oliver, N., Rosario, B. & Pentland, A.	A Bayesian Computer Vision System for Modeling Human Interactions [Abstract] [BibTeX]	2000	IEEE Transactions on Pattern Analysis and Machine Intelligence	article	URL
Abstract: We describe a real-time computer vision and machine learning system for modeling and recognizing human behaviors in a visual surveillance task [1]. The system is particularly concerned with detecting when interactions between people occur and classifying the type of interaction. Examples of interesting interaction behaviors include following another person, altering one's path to meet another, and so forth. Our system combines top-down with bottom-up information in a closed feedback loop, with both components employing a statistical Bayesian approach [2]. We propose and compare two different state-based learning architectures, namely, HMMs and CHMMs for modeling behaviors and interactions. The CHMM model is shown to work much more efficiently and accurately. Finally, to deal with the problem of limited training data, a synthetic “Alife-style” training system is used to develop flexible prior models for recognizing human interactions. We demonstrate the ability to use these a priori models to accurately classify real human behaviors and interactions with no additional tuning or training.
BibTeX: @article{Oliver2000, author = {Oliver, Nuria and Rosario, Barbara and Pentland, Alex}, title = {A Bayesian Computer Vision System for Modeling Human Interactions}, journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence}, year = {2000}, volume = {22}, number = {8}, pages = {831-843}, url = {citeseer.ist.psu.edu/article/oliver99bayesian.html} }
Ono, Y., Okabe, T. & Sato, Y.	Gaze Estimation from Low Resolution Images [BibTeX]	2006	LNCS: Advances in Image and Video Technology	inproceedings	URL
BibTeX: @inproceedings{Ono2006, author = {Ono, Y. and Okabe, T. and Sato, Y.}, title = {Gaze Estimation from Low Resolution Images}, booktitle = {LNCS: Advances in Image and Video Technology}, year = {2006}, pages = {178-188}, url = {http://www.springerlink.com/content/ru36223222615752/} }
Owechko, Y., Medasani, S. & Srinivasa, N.	Classifier Swarms for Human Detection in Infrared Imagery [Abstract] [BibTeX]	2004	OTCBVS	conference
Abstract: In this paper, we describe a new method for visual recognition of objects in an image that combines feature-based object classification with efficient search mechanisms based on swarm intelligence. Our approach utilizes the particle swarm optimization algorithm (PSO), a population based evolutionary algorithm, which is effective for optimization of a wide range of functions. PSO searches a multi-dimensional solution space for a global optimum using a population of "particles" in which each particle has its own velocity vector. In our approach, we extend PSO using sequential niching methods to handle multiple minima. Also, in our approach, each particle in the swarm is actually a self-contained classifier that "flys" through the solution space seeking the most "object-like" regions. By performing this optimization, the classifier swarm simultaneously finds objects in the scene, determines their size, and optimizes the classifier parameters.
BibTeX: @conference{Owechko2004, author = {Owechko, Yuri and Medasani, Swarup and Srinivasa, Narayan}, title = {Classifier Swarms for Human Detection in Infrared Imagery}, booktitle = {OTCBVS}, year = {2004} }
Pauwels, E. J. & Frederix, G.	Finding salient regions in images: nonparametric clustering for image segmentation and grouping [BibTeX]	1999	Comput. Vis. Image Underst.	article	DOI
BibTeX: @article{Pauwels1999, author = {E. J. Pauwels and G. Frederix}, title = {Finding salient regions in images: nonparametric clustering for image segmentation and grouping}, journal = {Comput. Vis. Image Underst.}, publisher = {Elsevier Science Inc.}, year = {1999}, volume = {75}, number = {1-2}, pages = {73--85}, doi = {http://dx.doi.org/10.1006/cviu.1999.0763} }
Pennec, X.	Computing the Mean of Geometric Features Application to the Mean Rotation [Abstract] [BibTeX]	1998		techreport	URL
Abstract: The question we investigate in this article is: what is the mean value of a set of geometric features and how can we compute it? We use as a guiding example one of the most studied type of features in computer vision and robotics: 3D rotations. The usual techniques on points consist of minimizing the least-square criterion, which gives the barycenter, the weighted least-squares or the sum of (squared) Mahalanobis distances. Unfortunately, these techniques rely on the vector space structure of points and generalizing them directly to other types of features could lead to paradoxes tePennec:JMIV:97. For instance, computing the barycenter of rotations using rotation matrices, unit quaternions or rotation vectors gives three different results. We present in this article a thorough generalization of the three above criterions to homogeneous Riemannian manifolds that rely only on intrinsic characteristics of the manifold. The necessary condition for the mean rotation, independently derived in tedenney96, is obtained here as a particular case of a general formula. We also propose an intrinsic gradient descent algorithm to obtain the minimum of the criterions and show how to estimate the uncertainty of the resulting estimation. These algorithms prove to be not only accurate but also efficient: computations are only 3 to 4 times longer for rotations than for points. The accuracy prediction of the results is within 1%, which is quite remarkable. The striking similarity of the algorithms' behavior for general features and for points stresses the validity of our approach using Riemannian geometry and lets us anticipate that other statistical results and algorithms could be generalized to manifolds in this framework.
BibTeX: @techreport{Pennec1998, author = {Pennec, Xavier}, title = {Computing the Mean of Geometric Features Application to the Mean Rotation}, year = {1998}, number = {3371}, url = {http://www.inria.fr/rrrt/rr-3371.html} }
Peters, R. J. & Itti, L.	Beyond bottom-up: Incorporating task-dependent influences into a computational model of spatial attention [Abstract] [Review] [BibTeX]	2007	Proc. IEEE Conference on Computer Vision and Pattern Recognition (CVPR)	inproceedings
Abstract: A critical function in both machine vision and biological vision systems is attentional selection of scene regions worthy of further analysis by higher-level processes such as object recognition. Here we present the first model of spatial attention that (1) can be applied to arbitrary static and dynamic image sequences with interactive tasks and (2) combines a general computational implementation of both bottom-up (BU) saliency and dynamic top-down (TD) task relevance; the claimed novelty lies in the combination of these elements and in the fully computational nature of the model. The BU component computes a saliency map from 12 low-level multi-scale visual features. The TD component computes a low-level signature of the entire image, and learns to associate different classes of signatures with the different gaze patterns recorded from human subjects performing a task of interest. We measured the ability of this model to predict the eye movements of people playing contemporary video games. We found that the TD model alone predicts where humans look about twice as well as does the BU model alone; in addition, a combined BU*TD model performs significantly better than either individual component. Qualitatively, the combined model predicts some easy-to-describe but hard-to-compute aspects of attentional selection, such as shifting attention leftward when approaching a left turn along a racing track. Thus, our study demonstrates the advantages of integrating BU factors derived from a saliency map and TD factors learned from image and task contexts in predicting where humans look while performing complex visually-guided behavior.
Review: full/conf
BibTeX: @inproceedings{Peters2007, author = {R. J. Peters and L. Itti}, title = {Beyond bottom-up: Incorporating task-dependent influences into a computational model of spatial attention}, booktitle = {Proc. IEEE Conference on Computer Vision and Pattern Recognition (CVPR)}, year = {2007} }
Philomin, V., Duraiswami, R. & Davis, L.	Quasi-Random Sampling for Condensation [Abstract] [BibTeX]	2000	European Conference on Computer Vision	inproceedings
Abstract: The problem of tracking pedestrians from a moving car isa challenging one. The Condensation tracking algorithm is appealingfor its generality and potential for real-time implementation. However,the conventional Condensation tracker is known to have diculty withhigh-dimensional state spaces and unknown motion models. This paperpresents an improved algorithm that addresses these problems by using asimplified motion model, and employing quasi-Monte Carlo techniques toeciently sample the resulting tracking problem in the high-dimensionalstate space. For N sample points, these techniques achieve sampling errorsof O(N1), as opposed to O(N1/2) for conventional Monte Carlotechniques. We illustrate the algorithm by tracking objects in both syntheticand real sequences, and show that it achieves reliable tracking andsignificant speed-ups over conventional Monte Carlo techniques.
BibTeX: @inproceedings{Philomin2000, author = {Philomin, Vasanth and Duraiswami, Ramani and Davis, Larry}, title = {Quasi-Random Sampling for Condensation}, booktitle = {European Conference on Computer Vision}, year = {2000}, pages = {--} }
Pieters, R. & Warlop, L.	Visual attention during brand choice : The impact of time pressure and task motivation [BibTeX]	1999	International Journal of Research in Marketing	article
BibTeX: @article{Pieters1999, author = {Pieters, R. and Warlop, L.}, title = {Visual attention during brand choice : The impact of time pressure and task motivation}, journal = {International Journal of Research in Marketing}, year = {1999}, volume = {16}, pages = {1-16} }
Pieters, R., Warlop, L. & Wedel, M.	Breaking Through the Clutter: Benefits of Advertisement Originality and Familiarity for Brand Attention and Memory [BibTeX]	2002	Manage. Sci.	article	DOI
BibTeX: @article{Pieters2002, author = {Rik Pieters and Luk Warlop and Michel Wedel}, title = {Breaking Through the Clutter: Benefits of Advertisement Originality and Familiarity for Brand Attention and Memory}, journal = {Manage. Sci.}, publisher = {INFORMS}, year = {2002}, volume = {48}, number = {6}, pages = {765--781}, doi = {http://dx.doi.org/10.1287/mnsc.48.6.765.192} }
Ramamoorthi, R., Koudelka, M. & Belhumeur, P.	A Fourier Theory for Cast Shadows [Abstract] [BibTeX]	2005	IEEE Transactions on Pattern Analysis and Machine Intelligence	article
Abstract: Cast shadows can be significant in many computer vision applications,such as lighting-insensitive recognition and surface reconstruction. Nevertheless,most algorithms neglect them, primarily because they involve nonlocal interactionsin nonconvex regions, making formal analysis difficult. However, many realinstances map closely to canonical configurations like a wall, a V-groove typestructure, or a pitted surface. In particular, we experiment with 3D textures likemoss, gravel, and a kitchen sponge, whose surfaces include canonicalconfigurations like V-grooves. This paper takes a first step toward a formalanalysis of cast shadows, showing theoretically that many configurations can bemathematically analyzed using convolutions and Fourier basis functions. Ouranalysis exposes the mathematical convolution structure of cast shadows andshows strong connections to recent signal-processing frameworks for reflectionand illumination.
BibTeX: @article{Ramamoorthi2005, author = {Ramamoorthi, Ravi and Koudelka, Melissa and Belhumeur, Peter}, title = {A Fourier Theory for Cast Shadows}, journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence}, year = {2005}, volume = {27}, number = {2}, pages = {--} }
Rao, C., Yilmaz, A. & Shah, M.	View-Invariant Representation and Recognition of Actions [BibTeX]	2002	International Journal of Computer Vision	article	DOI URL
BibTeX: @article{Rao2002, author = {Rao, Cen and Yilmaz, Alper and Shah, Mubarak}, title = {View-Invariant Representation and Recognition of Actions}, journal = {International Journal of Computer Vision}, publisher = {Kluwer Academic Publishers}, year = {2002}, volume = {50}, number = {2}, pages = {203--226}, url = {http://www.cs.ucf.edu/~vision/projects/ViewInvariance/ViewInvariance.html}, doi = {http://dx.doi.org/10.1023/A:1020350100748} }
Rao, R. P., Zelinsky, G. J., Hayhoe, M. M. & Ballard, D. H.	Eye movements in iconic visual search [BibTeX]	2002	Vision Research	article	DOI URL
BibTeX: @article{Rao2002a, author = {Rao, Rajesh P. and Zelinsky, Gregory J. and Hayhoe, Mary M. and Ballard, Dana H. }, title = {Eye movements in iconic visual search}, journal = {Vision Research}, year = {2002}, volume = {42}, number = {11}, pages = {1447--1463}, url = {http://dx.doi.org/10.1016/S0042-6989(02)00040-8}, doi = {http://dx.doi.org/10.1016/S0042-6989(02)00040-8} }
Rasmussen, C.	Grouping dominant orientations for ill-structured road following [Abstract] [BibTeX]	2004	International Conference on Computer Vision and Pattern Recognition	inproceedings
Abstract: Many rural roads lack sharp, smoothly curving edgesand a homogeneous surface appearance, hampering traditionalvision-based road-following methods. However,they often have strong texture cues parallel to the roaddirection in the form of ruts and tracks left by other vehicles.In this paper, we describe an algorithm for followingill-structured roads in which dominant textureorientations computed with multi-scale Gabor waveletfilters vote for a consensus road vanishing point location.In-plane road curvature and out-of-plane undulationare estimated in each image by tracking the vanishingpoint indicated by a horizontal image strip asit moves up toward the putative vanishing line. Particlefiltering is also used to track the vanishing pointsequence induced by road curvature from image to image.Results are shown for vanishing point localizationon a variety of road scenes ranging from gravel roadsto dirt trails to highways.
BibTeX: @inproceedings{Rasmussen2004, author = {Rasmussen, Christopher}, title = {Grouping dominant orientations for ill-structured road following}, booktitle = {International Conference on Computer Vision and Pattern Recognition}, year = {2004}, pages = {--} }
Rasmussen, C.	Texture-Based Vanishing Point Voting for Road Shape Estimation [Abstract] [BibTeX]	2004	British Machine Vision Conference	inproceedings
Abstract: Many rural roads lack sharp, smoothly curving edges and a homogeneous surfaceappearance, hampering traditional vision-based road-following methods. However,they often have strong texture cues parallel to the road direction in the form ofruts and tracks left by other vehicles. This paper describes an unsupervised algorithmfor following ill-structured roads in which dominant texture orientationscomputed with Gabor wavelet filters vote for a consensus road vanishing point location.The technique is first described for estimating the direction of straight-roadsegments, then extended to curved and undulating roads by tracking the vanishingpoint indicated by a dierential "strip" of voters moving up toward the nominalvanishing line. Finally, the vanishing point is used to constrain a search for theroad boundaries by maximizing texture- and color-based region discriminant functions.Results are shown for a variety of road scenes including gravel roads, dirttrails, and highways.
BibTeX: @inproceedings{Rasmussen2004a, author = {Rasmussen, Christopher}, title = {Texture-Based Vanishing Point Voting for Road Shape Estimation}, booktitle = {British Machine Vision Conference}, year = {2004}, pages = {--} }
Reynolds, C.	Steering Behaviors For Autonomous Characters [Abstract] [BibTeX]	1999	Game Developers Conference	conference
Abstract: This paper presents solutions for one requirement of autonomous characters in animation and games: the ability to navigate around their world in a life-like and improvisational manner. These "steering behaviors" are largely independent of the particulars of the character's means of locomotion. Combinations of steering behaviors can be used to achieve higher level goals (For example: get from here to there while avoiding obstacles, follow this corridor, join that group of characters...) This paper divides motion behavior into three levels. It will focus on the middle level of steering behaviors, briefly describe the lower level of locomotion, and touch lightly on the higher level of goal setting and strategy. Keywords: Animation Techniques, Virtual/Interactive Environments, Games, Simulation, behavioral animation, autonomous agent, situated, embodied, reactive, vehicle, steering, path planning, path following, pursuit, evasion, obstacle avoidance, collision avoidance, flocking, group behavior, navigation, artificial life, improvisation.
BibTeX: @conference{Reynolds1999, author = {Reynolds, Craig}, title = {Steering Behaviors For Autonomous Characters}, booktitle = {Game Developers Conference}, year = {1999} }
Reynolds, C.	Flocks, Herds, and Schools: A Distributed Behavioral Model [Abstract] [BibTeX]	1987	SIGGRAPH 87	conference	DOI
Abstract: The aggregate motion of a flock of birds, a herd of land animals, or a school of fish is a beautiful and familiar part of the natural world. But this type of complex motion is rarely seen in computer animation. This paper explores an approach based on simulation as an alternative to scripting the paths of each bird individually. The simulated flock is an elaboration of a particle system, with the simulated birds being the particles. The aggregate motion of the simulated flock is created by a distributed behavioral model much like that at work in a natural flock; the birds choose their own course. Each simulated bird is implemented as an independent actor that navigates according to its local perception of the dynamic environment, the laws of simulated physics that rule its motion, and a set of behaviors programmed into it by the "animator." The aggregate motion of the simulated flock is the result of the dense interaction of the relatively simple behaviors of the individual simulated birds.
BibTeX: @conference{Reynolds1987, author = {Reynolds, Craig}, title = {Flocks, Herds, and Schools: A Distributed Behavioral Model}, booktitle = {SIGGRAPH 87}, publisher = {ACM Press}, year = {1987}, pages = {25--34}, doi = {http://doi.acm.org/10.1145/37401.37406} }
Rittscher, J., Tu, P. & Krahnstoever, N.	Simultaneous Estimation of Segmentation and Shape [BibTeX]	2005	International Conference on Computer Vision and Pattern Recognition	inproceedings
BibTeX: @inproceedings{Rittscher2005, author = {Rittscher, J. and Tu, P.H. and Krahnstoever, N.}, title = {Simultaneous Estimation of Segmentation and Shape}, booktitle = {International Conference on Computer Vision and Pattern Recognition}, year = {2005}, pages = {II: 486-493} }
Roberts, T., McKenna, S. & Ricketts, I.	Adaptive Learning of Statistical Appearance Models for3D Human Tracking [Abstract] [BibTeX]	2002	British Machine Vision Conference	conference
Abstract: A likelihood formulation for human tracking is presented based upon matching feature statistics on the surface of an articulated 3D body model. A benefit of such a formulation over current techniques is that it provides a dense, object-based cue. Multi-dimensional histograms are used to represent feature distributions and different histogram similarity measures are evaluated. An on-line region grouping algorithm, driven by prior knowledge of clothing structure, is derived that enables better histogram estimation and greatly increases computational efficiency. Finally, we demonstrate that the smooth, broad likelihood response allows efficient inference using coarse sampling and local optimisation. Results from tracking real world sequences are presented. 1 Introduction
BibTeX: @conference{Roberts2002, author = {Roberts, Timothy and McKenna, Stephen and Ricketts, Ian}, title = {Adaptive Learning of Statistical Appearance Models for3D Human Tracking}, booktitle = {British Machine Vision Conference}, year = {2002} }
Robertson, N. & Reid, I.	Estimating Gaze Direction from Low-Resolution Faces in Video [BibTeX]	2006	ECCV	inproceedings	URL
BibTeX: @inproceedings{Robertson2006, author = {Neil Robertson and Ian Reid}, title = {Estimating Gaze Direction from Low-Resolution Faces in Video}, booktitle = {ECCV}, year = {2006}, pages = {402-415}, url = {http://www.springerlink.com/content/q2x2618252558352/} }
Robertson, N., Reid, I. & Brady, J.	What are you looking at? Gaze estimation in medium-scale images [Abstract] [BibTeX]	2005	HAREM	inproceedings
Abstract: In this paper we describe a new method for estimating where a person is looking in images where the head of a person is typically 20 pixels high. We use a feature vector based on skin detection to estimate the orientation of the head, which is discretised into 8 different orientations, relative to the camera. A fast sampling method returns a distribution over head pose. The generaldirectionofthepersonisestimatedbasedonvelocity. Weshowthat, by combining direction and head pose using a Bayesian Network gaze is determined more robustly than using each feature alone. We demonstrate this technique on surveillance and sportsfootage.
BibTeX: @inproceedings{Robertson2005, author = {Robertson, N.M. and Reid, I. and Brady, J.M.}, title = {What are you looking at? Gaze estimation in medium-scale images}, booktitle = {HAREM}, year = {2005} }
Romero, M. & Bobick, A.	Tracking Head Yaw by Interpolation of Template Responses [BibTeX]	2004	CVPRW '04: Proceedings of the 2004 Conference on Computer Vision and Pattern Recognition Workshop (CVPRW'04) Volume 5	inproceedings
BibTeX: @inproceedings{Romero2004, author = {Mario Romero and Aaron Bobick}, title = {Tracking Head Yaw by Interpolation of Template Responses}, booktitle = {CVPRW '04: Proceedings of the 2004 Conference on Computer Vision and Pattern Recognition Workshop (CVPRW'04) Volume 5}, publisher = {IEEE Computer Society}, year = {2004}, pages = {83} }
Rosenholtz, R., Li, Y. & Nakano, L.	Measuring visual clutter [Abstract] [BibTeX]	2007	J. Vis.	article	URL
Abstract: Visual clutter concerns designers of user interfaces and information visualizations. This should not surprise visual perception researchers because excess and/or disorganized display items can cause crowding, masking, decreased recognition performance due to occlusion, greater difficulty at both segmenting a scene and performing visual search, and so on. Given a reliable measure of the visual clutter in a display, designers could optimize display clutter. Furthermore, a measure of visual clutter could help generalize models like Guided Search (J. M. Wolfe, 1994) by providing a substitute for “set size” more easily computable on more complex and natural imagery. In this article, we present and test several measures of visual clutter, which operate on arbitrary images as input. The first is a new version of the Feature Congestion measure of visual clutter presented in R. Rosenholtz, Y. Li, S. Mansfield, and Z. Jin (2005). This Feature Congestion measure of visual clutter is based on the analogy that the more cluttered a display or scene is, the more difficult it would be to add a new item that would reliably draw attention. A second measure of visual clutter, Subband Entropy, is based on the notion that clutter is related to the visual information in the display. Finally, we test a third measure, Edge Density, used by M. L. Mack and A. Oliva (2004) as a measure of subjective visual complexity. We explore the use of these measures as stand-ins for set size in visual search models and demonstrate that they correlate well with search performance in complex imagery. This includes the search-in-clutter displays of J. M. Wolfe, A. Oliva, T. S. Horowitz, S. Butcher, and A. Bompas (2002) and Bravo and Farid (2004), as well as new search experiments. An additional experiment suggests that color variability, accounted for by Feature Congestion but not the Edge Density measure or the Subband Entropy measure, does matter for visual clutter.
BibTeX: @article{Rosenholtz2007, author = {Rosenholtz, Ruth and Li, Yuanzhen and Nakano, Lisa}, title = {Measuring visual clutter}, journal = {J. Vis.}, year = {2007}, volume = {7}, number = {2}, pages = {1-22}, note = {Matlab code downloadable at http://dspace.mit.edu/handle/1721.1/37593}, url = {http://journalofvision.org/7/2/17/} }
Roth, D., Doubek, P. & Van Gool, L.	Bayesian Pixel Classification for Human Tracking [BibTeX]	2005	Workshop on Motion and Video Computing	inproceedings
BibTeX: @inproceedings{Roth2005, author = {Roth, Daniel and Doubek, Petr and Van Gool, Luc}, title = {Bayesian Pixel Classification for Human Tracking}, booktitle = {Workshop on Motion and Video Computing}, publisher = {IEEE}, year = {2005}, pages = {78-83} }
Sebe, N., Cohen, I., Huang, T. S. & Gevers, T.	Skin Detection: A Bayesian Network Approach [Abstract] [BibTeX]	2004	International Conference on Pattern Recognition	conference	URL
Abstract: The automated detection and tracking of humans in computer vision necessitates improved modeling of the human skin appearance. In this paper we propose a Bayesian network approach for skin detection. We test several classifiers and propose a methodology for incorporating unlabeled data. We apply the semi-supervised approach to skin detection and we show that learning the structure of Bayesian network classifiers enables learning good classifiers with a small labeled set and a large unlabeled set.
BibTeX: @conference{Sebe2004, author = {Sebe, Nicu and Cohen, Ira and Huang, Thomas S. and Gevers, Theo}, title = {Skin Detection: A Bayesian Network Approach}, booktitle = {International Conference on Pattern Recognition}, year = {2004}, url = {http://citeseer.ifi.unizh.ch/665466.html} }
Shahrokni, A., Drummond, T. & Fua, P.	Texture boundary detection for real-time tracking [Abstract] [BibTeX]	2004	European Conference on Computer Vision	inproceedings
Abstract: We propose an approach to texture boundary detection that only requires a line-search in the direction normal to the edge. It is therefore very fast and can be incorporated into a real-time 3-D pose estimation algorithm that retains the speed of those that rely solely on gradient properties along object contours but does not fail in the presence of highly textured object and clutter. This is achieved by correctly integrating probabilities over the space of statistical texture models. We will show that this rigorous and formal statistical treatment results in good performance under demanding circumstances
BibTeX: @inproceedings{Shahrokni2004a, author = {Shahrokni, Ali and Drummond, Tom and Fua, P.}, title = {Texture boundary detection for real-time tracking}, booktitle = {European Conference on Computer Vision}, year = {2004}, volume = {3022}, pages = {566--577} }
Shahrokni, A., Lepetit, V., Drummond, T. & Fua, P.	Markov-based Silhouette Extraction for ThreenDimensional Body Tracking in Presence of Cluttered Background [Abstract] [BibTeX]	2004	British Machine Vision Conference	inproceedings	URL
Abstract: We propose a novel method to detect human body contours in presence ofclutter and complex texture. Contours are extracted using a novel Markovbasedapproach which learns a texture along a given scanline in order todetect texture crossings. In contrast to conventional silhouette detection algorithmsbased on gradient, our texture boundary detection method allowsextraction of silhouettes of textured and non-textured objects under dif-cult conditions such as having a cluttered/moving background. We demonstrateon demanding examples of monocular body tracking that our proposedmethod yields better results than gradient-based techniques.
BibTeX: @inproceedings{Shahrokni2004, author = {Shahrokni, Ali and Lepetit, Vincent and Drummond, Tom and Fua, Pascal}, title = {Markov-based Silhouette Extraction for ThreenDimensional Body Tracking in Presence of Cluttered Background}, booktitle = {British Machine Vision Conference}, year = {2004}, pages = {--}, url = {http://ligwww.epfl.ch/~fua/} }
Sheikh, Y. & Shah, M.	Bayesian Object Detection in Dynamic Scenes [Abstract] [BibTeX]	2005	International Conference on Computer Vision and Pattern Recognition	inproceedings
Abstract: Detecting moving objects using stationary cameras is an important precursor to many activity recognition, object recognition and tracking algorithms. In this paper, three innovations are presented over existing approaches. Firstly, the model of the intensities of image pixels as independently distributed random variables is challenged and it is asserted that useful correlation exists in the intensities of spatially proximal pixels. This correlation is exploited to sustain high levels of detection accuracy in the presence of nominal camera motion and dynamic textures. By using a non-parametric density estimation method over a joint domain-range representation of image pixels, multimodal spatial uncertainties and complex dependencies between the domain (location) and range (color) are directly modeled. Secondly, temporal persistence is proposed as a detection criteria. Unlike previous approaches to object detection which detect objects by building adaptive models of the only background, the foreground is also modeled to augment the detection of objects (without explicit tracking) since objects detected in a preceding frame contain substantial evidence for detection in a current frame. Third, the background and foreground models are used competitively in a MAP-MRF decision framework, stressing spatial context as a condition of pixel-wise labeling and the posterior function is maximized efficiently using graph cuts. Experimental validation of the proposed method is presented on a diverse set of dynamic scenes.
BibTeX: @inproceedings{Sheikh2005, author = {Sheikh, Yaser and Shah, Mubarak}, title = {Bayesian Object Detection in Dynamic Scenes}, booktitle = {International Conference on Computer Vision and Pattern Recognition}, year = {2005} }
Sidenbladh, H., Black, M. & Fleet, D.	Stochastic Tracking of 3D Human Figures Using 2D Image Motion [Abstract] [BibTeX]	2000	European Conference on Computer Vision	inproceedings
Abstract: A probabilistic method for tracking 3D articulated human figures in monocular image sequences is presented. Within a Bayesian framework, we define a generative model of image appearance, a robust likelihood function based on image graylevel differences, and a prior probability distribution over pose and joint angles that models how humans move. The posterior probability distribution over model parameters is represented using a discrete set of samples and is propagated over time using particle filtering. The approach extends previous work on parameterized optical flow estimation to exploit a complex 3D articulated motion model. It also extends previous work on human motion tracking by including a perspective camera model, by modeling limb self occlusion, and by recovering 3D motion from a monocular sequence. The explicit posterior probability distribution represents ambiguities due to image matching, model singularities, and perspective projection. The method relies only on a frame-to-frame assumption of brightness constancy and hence is able to track people under changing viewpoints, in grayscale image sequences, and with complex unknown backgrounds.
BibTeX: @inproceedings{Sidenbladh2000, author = {Sidenbladh, Hedvig and Black, Michael and Fleet, David}, title = {Stochastic Tracking of 3D Human Figures Using 2D Image Motion}, booktitle = {European Conference on Computer Vision}, year = {2000}, pages = {--} }
Sidenbladh, H., Black, M. & Sigal, L.	Implicit probabilistic models of human motion for synthesis and tracking [Abstract] [BibTeX]	2002	European Conference on Computer Vision	inproceedings
Abstract: This paper addresses the problem of probabilistically modeling 3D human motion for synthesis and tracking. Given the high dimensional nature of human motion, learning an explicit probabilistic model from available training data is currently impractical. Instead we exploit methods from texture synthesis that treat images as representing an implicit empirical distribution. These methods replace the problem of representing the probability of a texture pattern with that of searching the training data for similar instances of that pattern. We extend this idea to temporal data representing 3D human motion with a large database of example motions. To make the method useful in practice, we must address the problem of efficient search in a large training set; efficiency is particularly important for tracking. Towards that end, we learn a low dimensional linear model of human motion that is used to structure the example motion database into a binary tree. An approximate probabilistic tree search method exploits the coefficients of this low-dimensional representation and runs in sub-linear time. This probabilistic tree search returns a particular sample human motion with probability approximating the true distribution of human motions in the database. This sampling method is suitable for use with particle filtering techniques and is applied to articulated 3D tracking of humans within a Bayesian framework. Successful tracking results are presented, along with examples of synthesizing human motion using the model.
BibTeX: @inproceedings{Sidenbladh2002, author = {Sidenbladh, Hedvig and Black, Michael and Sigal, Leonid}, title = {Implicit probabilistic models of human motion for synthesis and tracking}, booktitle = {European Conference on Computer Vision}, year = {2002}, volume = {1}, pages = {784--800} }
Sminchisescu, C. & Triggs, B.	Kinematic Jump Processes For Monocular 3D Human Tracking [Abstract] [BibTeX]	2003	International Conference on Computer Vision and Pattern Recognition	inproceedings	DOI URL
Abstract: A major difficulty for 3D human body tracking from monocular image sequences is the near non-observability of kinematic degrees of freedom that generate motion in depth. For known link (body seg-ment) lengths, the strict non-observabilities reduce to twofold ‘for-wards/ backwards flipping’ ambiguities for each link. These imply 2# links formal inverse kinematics solutions for the full model, and hence linked groups of O(2# links) local minima in the model-image matching cost function. Choosing the wrong minimum leads to rapid mistracking, so for reliable tracking, rapid methods of inestigating alternative minima within a group are needed. Previous approaches to this have used generic search methods that do not exploit the specific problem structure. Here, we complement these by using simple kinematic reasoning to enumerate the tree of possible forwards/backwards flips, thus greatly speeding the search within each linked group of minima. Our methods can be used either deterministically, or within stochastic ‘jump-diffusion’ style search processes. We give experimental results on some challenging monocular human tracking sequences, showing how the new kinematic-flipping based sampling method improves and complements existing ones.
BibTeX: @inproceedings{Sminchisescu2003, author = {Sminchisescu, Cristian and Triggs, Bill}, title = {Kinematic Jump Processes For Monocular 3D Human Tracking}, booktitle = {International Conference on Computer Vision and Pattern Recognition}, publisher = {IEEE Computer Society}, year = {2003}, volume = {01}, pages = {69}, url = {http://doi.ieeecomputersociety.org/10.1109/CVPR.2003.1211339}, doi = {http://doi.ieeecomputersociety.org/10.1109/CVPR.2003.1211339} }
Smith, K., Gatica-Perez, D. & Odobez, J.	Using Particles to Track Varying Numbers of Interacting People [Abstract] [BibTeX]	2005	International Conference on Computer Vision and Pattern Recognition	inproceedings	URL
Abstract: In this paper, we present a Bayesian framework for the fully automatic tracking of a variable number of interacting targets using a fixed camera. This framework uses a joint multi-object state-space formulation and a transdimensional Markov Chain Monte Carlo (MCMC) particle filter to recursively estimate the multi-object configuration and efficiently search the state-space. We also define a global observation model comprised of color and binary measurements capable of discriminating between different numbers of objects in the scene. We present results which show that our method is capable of tracking varying numbers of people through several challenging real-world tracking situations such as full/partial occlusion and entering/ leaving the scene.
BibTeX: @inproceedings{Smith2005, author = {Smith, Kevin and Gatica-Perez, Daniel and Odobez, Jean-Marc}, title = {Using Particles to Track Varying Numbers of Interacting People}, booktitle = {International Conference on Computer Vision and Pattern Recognition}, year = {2005}, url = {http://www.idiap.ch/~gatica/publications/pub.html} }
Starck, J. & Hilton, A.	Model-Based Multiple View Reconstruction of People [BibTeX]	2003	International Conference on Computer Vision	conference	DOI
BibTeX: @conference{Starck2003, author = {Jonathan Starck and Adrian Hilton}, title = {Model-Based Multiple View Reconstruction of People}, booktitle = {International Conference on Computer Vision}, publisher = {IEEE Computer Society}, year = {2003}, volume = {02}, pages = {915}, doi = {http://doi.ieeecomputersociety.org/10.1109/ICCV.2003.1238446} }
Stauffer, C. & Grimson, W.	Adaptive background mixture models for real-time tracking [Abstract] [BibTeX]	1999	International Conference on Computer Vision and Pattern Recognition	inproceedings
Abstract: A common method for real-time segmentation ofmoving regions in image sequences involves "backgroundsubtraction," or thresholding the error betweenan estimate of the image without moving objects andthe current image. The numerous approaches to thisproblem differ in the type of background model usedand the procedure used to update the model. This paperdiscusses modeling each pixel as a mixture of Gaussiansand using an on-line approximation to updatethe model. The Gaussian distributions of the adaptivemixture model are then evaluated to determine whichare most likelyt o result from a background process.Each pixel is classified based on whether the Gaussiandistribution which represents it most effectivelyis consideredpart of the background model.This results in a stable, real-time outdoor trackerwhich reliablyde als with lighting changes, repetitivemotions from clutter, and long-term scene changes.This system has been run almost continuously for 16months, 24 hours a day, through rain and snow.
BibTeX: @inproceedings{Stauffer1999, author = {Stauffer, Chris and Grimson, W.}, title = {Adaptive background mixture models for real-time tracking}, booktitle = {International Conference on Computer Vision and Pattern Recognition}, year = {1999}, volume = {2}, pages = {252} }
Strom, J., Jebara, T., Basu, S. & Pentland, A.	Real Time Tracking and Modeling of Faces: An EKF-Based Analysis by Synthesis Approach [BibTeX]	1999	MPEOPLE '99: Proceedings of the IEEE International Workshop on Modelling People	inproceedings
BibTeX: @inproceedings{Strom1999, author = {J. Strom and T. Jebara and S. Basu and A. Pentland}, title = {Real Time Tracking and Modeling of Faces: An EKF-Based Analysis by Synthesis Approach}, booktitle = {MPEOPLE '99: Proceedings of the IEEE International Workshop on Modelling People}, publisher = {IEEE Computer Society}, year = {1999}, pages = {55} }
Tabb, K., Davey, N., Adams, R. & George, S.	Detecting, Tracking & Classifying Human Movement using Active Contour Models and Neural Networks [BibTeX]	2004		inbook	URL
BibTeX: @inbook{Tabb2004, author = {Tabb, Ken and Davey, Neil and Adams, Rod and George, Stella}, title = {Detecting, Tracking & Classifying Human Movement using Active Contour Models and Neural Networks}, year = {2004}, url = {http://www.health.herts.ac.uk/ken/vision/} }
Tabb, K., Davey, N., Adams, R. & George, S.	A Hybrid Detection and Classification System for Human Motion Analysis [Abstract] [BibTeX]	2002		inbook	URL
Abstract: This paper discusses a hybrid technique for detecting and tracking movingpedestrians in a video sequence. The technique comprises two sub-systems: anactive contour model for detecting and tracking moving objects in the visual field,and an MLP neural network for classifying the moving objects being tracked ashuman or non-human . The axis crossover vector method is used for translatingthe active contour into a scale-. location-, resolution- and rotation-invariantvector suited for input to a neural network, and we identify the most appropriatelevel of detail for encoding human shape information. Experiments measuring theneural network s accuracy at classifying unseen computer generated and realmoving objects are presented, along with potential applications of the technology.Previous work has accommodated lateral pedestrian movement across the visualfield; this paper describes a system which accommodates arbitrary angles ofpedestrian movement on the ground plane.
BibTeX: @inbook{Tabb2002, author = {Tabb, Ken and Davey, Neil and Adams, Rod and George, Stella}, title = {A Hybrid Detection and Classification System for Human Motion Analysis}, year = {2002}, pages = {139--151}, url = {http://www.health.herts.ac.uk/ken/vision/} }
Tao, H., Sawhney, H. S. & Kumar, R.	A Sampling Algorithm for Tracking Multiple Objects [BibTeX]	1999	Workshop on Vision Algorithms	inproceedings	URL
BibTeX: @inproceedings{Tao1999, author = {Hai Tao and Harpreet S. Sawhney and Rakesh Kumar}, title = {A Sampling Algorithm for Tracking Multiple Objects}, booktitle = {Workshop on Vision Algorithms}, year = {1999}, pages = {53-68}, url = {citeseer.ist.psu.edu/tao99sampling.html} }
Torralba, A.	Contextual influences on saliency [BibTeX]	2005	Neurobiology of attention	article
BibTeX: @article{Torralba2005, author = {Torralba, A. }, title = {Contextual influences on saliency}, journal = {Neurobiology of attention}, publisher = {Academic Press}, year = {2005}, pages = {586--592} }
Traver, V., Bernardino, A., Moreno, P. & Santos-Victor, J.	Appearance-based object detection in space-variant images: a multi-model approach [Abstract] [BibTeX]	2004	International Conference on Image Analysis and Recognition	inproceedings	URL
Abstract: Recently, log-polar images have been successfully used inactive-vision tasks such as vergence control or target tracking. However,while the role of foveal data has been exploited and is well known, thatof periphery seems underestimated and not well understood. Nevertheless,peripheral information becomes crucial in detecting non-foveatedobjects or events. In this paper, a multiple-model approach (MMA) fortop-down, model-based attention processes is proposed. The advantagesoffered by this proposal for space-variant image representations are discussed.A simple but representative frontal-face detection task is given asan example of application of the MMA. The combination of appearancebasedfeatures and a linear regression-based classifier proved very effective.Results show the ability of the system to detect faces at very lowresolutions, which has implications in fields such as visual surveillance.
BibTeX: @inproceedings{Traver2004, author = {Traver, V.J. and Bernardino, A and Moreno, P and Santos-Victor, J}, title = {Appearance-based object detection in space-variant images: a multi-model approach}, booktitle = {International Conference on Image Analysis and Recognition}, year = {2004}, pages = {--}, url = {http://homepages.inf.ed.ac.uk/rbf/CAVIAR/} }
Tsiamyrtzis, P., Dowdall, J., Shastri, D., Pavlidis, I. T., Frank, M. G. & Ekman, P.	Imaging Facial Physiology for the Detection of Deceit [BibTeX]	2007	International Journal of Computer Vision	article
BibTeX: @article{Tsiamyrtzis2007, author = {P. Tsiamyrtzis and J. Dowdall and D. Shastri and I. T. Pavlidis and M. G. Frank and Ekman,P.}, title = {Imaging Facial Physiology for the Detection of Deceit}, journal = {International Journal of Computer Vision}, year = {2007}, volume = {71 (2)}, pages = {197-214} }
Uyttendaele, M., Criminisi, A., Kang, S. B., Winder, S., Szeliski, R. & Hartley, R.	Image-Based Interactive Exploration of Real-World Environments [BibTeX]	2004	IEEE Computer Graphics and Applications	article	URL
BibTeX: @article{Uyttendaele2004, author = {Uyttendaele, Matthew and Criminisi,Antonio and Kang, Sing Bing and Winder, Simon and Szeliski, Richard and Hartley, Richard}, title = {Image-Based Interactive Exploration of Real-World Environments}, journal = {IEEE Computer Graphics and Applications}, year = {2004}, volume = {24}, number = {3}, pages = {52-63}, url = {http://research.microsoft.com/users/mattu/pubs/tr-2003-61.pdf} }
Wada, T. & Matsuyama, T.	Multiobject Behavior Recognition by Event Driven Selective Attention Method [BibTeX]	2000	IEEE Transactions on Pattern Analysis and Machine Intelligence	article	DOI
BibTeX: @article{Wada2000, author = {Toshikazu Wada and Takashi Matsuyama}, title = {Multiobject Behavior Recognition by Event Driven Selective Attention Method}, journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence}, publisher = {IEEE Computer Society}, year = {2000}, volume = {22}, number = {8}, pages = {873-887}, doi = {http://doi.ieeecomputersociety.org/10.1109/34.868687} }
Walther, D.	Interactions of visual attention and object recognition : computational modeling, algorithms, and psychophysics [BibTeX]	2006	School: Caltech	phdthesis	URL
BibTeX: @phdthesis{Walther2006a, author = {Walther, Dirk}, title = {Interactions of visual attention and object recognition : computational modeling, algorithms, and psychophysics}, school = {Caltech}, year = {2006}, url = {http://resolver.caltech.edu/CaltechETD:etd-03072006-135433} }
Walther, D. & Koch, C.	2006 Special Issue: Modeling attention to salient proto-objects [BibTeX]	2006	Neural Netw.	article	DOI
BibTeX: @article{Walther2006, author = {Dirk Walther and Christof Koch}, title = {2006 Special Issue: Modeling attention to salient proto-objects}, journal = {Neural Netw.}, publisher = {Elsevier Science Ltd.}, year = {2006}, volume = {19}, pages = {1395--1407}, doi = {http://dx.doi.org/10.1016/j.neunet.2006.10.001} }
Wang, G., Tsui, H., Hu, Z. & Wu, F.	Camera calibration and 3D reconstruction from a single view based on scene constraints [Abstract] [BibTeX]	2005	Image and Vision Computing	article	URL
Abstract: This paper mainly focuses on the problem of camera calibration and 3D reconstruction from a single view of structured scene. It is well known that three constraints on the intrinsic parameters of a camera can be obtained from the vanishing points of three mutually orthogonal directions. However, there usually exist one or several pairs of line segments, which are mutually orthogonal and lie in the pencil of planes defined by two of the vanishing directions in the structured scenes. It is proved in this paper that a new independent constraint to the image of the absolute conic can be obtained if the pair of line segments is of equal length or with known length ratio in space. The constraint is further studied both in terms of the vanishing points and the images of circular points. Hence, four independent constraints on a camera are obtained from one image, and the camera can be calibrated under the widely accepted assumption of zero-skew. This paper also presents a simple method for the recovery of camera extrinsic parameters and projection matrix with respect to a given world coordinate system. Furthermore, several methods are presented to estimate the positions and poses of space planar surfaces from the recovered projection matrix and scene constraints. Thus, a scene structure can be reconstructed by combining the planar patches. Extensive experiments on simulated data and real images, as well as a comparative test with other methods in the literature, validate our proposed methods. (C) 2004 Elsevier B.V. All rights reserved
BibTeX: @article{Wang2005, author = {Wang, Guanhui. and Tsui, Hung-Tat. and Hu, Zhanyi and Wu, Fuchao}, title = {Camera calibration and 3D reconstruction from a single view based on scene constraints}, journal = {Image and Vision Computing}, year = {2005}, volume = {23}, number = {3}, pages = {311--323}, url = {ISI:000227221800004} }
Wang, J. & Singh, S.	Video analysis of human dynamics: a survey [Abstract] [BibTeX]	2003	Real Time Imaging	article	URL
Abstract: Video analysis of human dynamics is an important area of research devoted to detecting people and understanding their dynamic physical behavior in a complex environment that can be used for biometric applications. This paper provides a detailed survey of the various studies in areas related to the tracking of people and body parts such as face, hands, fingers, legs, etc., and modeling behavior using motion analysis.
BibTeX: @article{Wang2003, author = {Wang, Jessica and Singh, Sameer}, title = {Video analysis of human dynamics: a survey}, journal = {Real Time Imaging}, year = {2003}, volume = {9}, number = {5}, pages = {321–346}, url = {http://www.dcs.ex.ac.uk/research/pann/master/publication_copy.html} }
Wang, J. Z., Li, J., Wiederhold, G. & Firschein, O.	System for Screening Objectionable Images [Abstract] [BibTeX]	1998	Computer Communications Journal	article	URL
Abstract: As computers and Internet become more and more available to families, access of objectionable graphics by children is increasingly a problem that many parents are concerned about. This paper describes WIPE(TM) (Wavelet Image Pornography Elimination), a system capable of classifying an image as objectionable or benign. The algorithm uses a combination of an icon filter, a graph-photo detector, a color histogram filter, a texture filter, and a wavelet-based shape matching algorithm to provide robust screening of on-line objectionable images. Semantically-meaningful feature vector matching is carried out so that comparisons between a given on-line image and images in a pre-marked training data set can be performed efficiently and effectively. The system is practical for real-world applications, processing queries at the speed of less than 2 seconds each, including the time to compute the feature vector for the query, on a Pentium Pro PC. Besides its exceptional speed, it has demonstrated 96% sensitivity over a test set of 1,076 digital photographs found on objectionable news groups. It wrongly classified 9% of a set of 10,809 benign photographs obtained from various sources. The specificity in real-world applications is expected to be much higher because benign on-line graphs can be filtered out with our graph-photo detector with 100% sensitivity and nearly 100% specificity, and surrounding text can be used to assist the classification process.
BibTeX: @article{Wang1998, author = {Wang, James Z. and Li, Jia and Wiederhold, Gio and Firschein, Oscar}, title = {System for Screening Objectionable Images}, journal = {Computer Communications Journal}, year = {1998}, url = {http://www-db.stanford.edu/IMAGE/JCC98/} }
Wilczkowiak, M., Boyer, E. & Sturm, P.	3D modelling using geometric constraints: a parallelepiped based approach [Abstract] [BibTeX]	2002		inproceedings	URL
Abstract: In this paper, efficient and generic tools for calibration and 3D reconstruction are presented. These tools exploit geometric constraints frequently present in man-made environments and allow camera calibration as well as scene structure to be estimated with a small amount of user interactions and little a priori knowledge. The proposed approach is based on primitives that naturally characterize rigidity constraints: parallelepipeds. It has been shown previously that the intrinsic metric characteristics of a parallelepiped are dual to the intrinsic characteristics of a perspective camera. Here, we generalize this idea by taking into account additional redundancies between multiple images of multiple parallelepipeds. We propose a method for the estimation of camera and scene parameters that bears strong similarities with some self-calibration approaches. Taking into account prior knowledge on scene primitives or cameras, leads to simpler equations than for standard self-calibration, and is expected to improve results, as well as to allow structure and motion recovery in situations that are otherwise under-constrained. These principles are illustrated by experimental calibration results and several reconstructions from uncalibrated images
BibTeX: @inproceedings{Wilczkowiak2002, author = {Wilczkowiak, Marta and Boyer, Edmond and Sturm, Peter}, title = {3D modelling using geometric constraints: a parallelepiped based approach}, year = {2002}, volume = {2353}, pages = {221--236}, url = {ISI:000180067300015} }
Wilczkowiak, M., Boyer, E. & Sturm, P.	Camera Calibration and 3D Reconstruction from Single Images Using Parallelepipeds [Abstract] [BibTeX]	2001	International Conference on Computer Vision	inproceedings
Abstract: In this paper, parallelepipeds and their use in camera calibration and 3D reconstruction processes are studied. Parallelepipeds naturally characterize rigidity constraints present in a scene, such as parallelism and orthogonality. A subclass of parallelepipeds - the cuboids - has been frequently used over the past to partially calibrate cameras. However, the full potential of parallelepipeds, in camera calibration as well as in scene reconstruction, has never been clearly established. We propose a new framework for the use of parallelepipeds which is based on an extensive study of this potential. In particular, we exhibit the complete duality that exists between the intrinsic metric characteristics of a parallelepiped and the intrinsic parameters of a camera. Our framework allows to fully exploit parallelepipeds and thus overcomes several limitations of calibration approaches based on cuboids. To illustrate this framework, we present an original and very efficient interactive method for 3D reconstruction from single images. This method allows to quickly build a scene model from a single uncalibrated image.
BibTeX: @inproceedings{Wilczkowiak2001, author = {Wilczkowiak, Marta and Boyer, Edmond and Sturm, Peter}, title = {Camera Calibration and 3D Reconstruction from Single Images Using Parallelepipeds}, booktitle = {International Conference on Computer Vision}, year = {2001}, pages = {--} }
Wiles, C., Maki, A., Matsuda, N. & Watanabe, M.	Hyper-patches for 3D model acquisition and tracking [BibTeX]	1997	cvpr	article	DOI
BibTeX: @article{Wiles1997, author = {C.S. Wiles and A. Maki and N. Matsuda and M. Watanabe}, title = {Hyper-patches for 3D model acquisition and tracking}, journal = {cvpr}, publisher = {IEEE Computer Society}, year = {1997}, volume = {00}, pages = {1074}, doi = {http://doi.ieeecomputersociety.org/10.1109/CVPR.1997.609463} }
Wooldridge, M.	An Introduction to MultiAgent Systems, 366 pages, ISBN 0-471-49691-X. [BibTeX]	2002		book	URL
BibTeX: @book{Wooldridge2002, author = {Wooldridge, Michael}, title = {An Introduction to MultiAgent Systems, 366 pages, ISBN 0-471-49691-X.}, publisher = {John Wiley & Sons Ltd}, year = {2002}, url = {http://www.csc.liv.ac.uk/~mjw/pubs/imas/} }
Wren, C. R., Azarbayejani, A., Darrell, T. & Pentland, A.	Pfinder: Real-Time Tracking of the Human Body [BibTeX]	1997	IEEE Transactions on Pattern Analysis and Machine Intelligence	article	URL
BibTeX: @article{Wren1997, author = {Christopher Richard Wren and Ali Azarbayejani and Trevor Darrell and Alex Pentland}, title = {Pfinder: Real-Time Tracking of the Human Body}, journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence}, year = {1997}, volume = {19}, number = {7}, pages = {780-785}, url = {citeseer.ist.psu.edu/wren97pfinder.html} }
Xiang, T. & Gong, S.	Beyond Tracking: Modelling Activity and Understanding Behaviour [BibTeX]	2006	Int. J. Comput. Vision	article	DOI
BibTeX: @article{1117998, author = {Tao Xiang and Shaogang Gong}, title = {Beyond Tracking: Modelling Activity and Understanding Behaviour}, journal = {Int. J. Comput. Vision}, publisher = {Kluwer Academic Publishers}, year = {2006}, volume = {67}, number = {1}, pages = {21--51}, doi = {http://dx.doi.org/10.1007/s11263-006-4329-6} }
Xie, L., Chang, S., Divakaran, A. & Sun, H.	Unsupervised discovery of multilevel statistical video structures using hierarchical hidden Markov models [BibTeX]	2003	IEEE Intl. Conf. Multimedia and Expo (ICME)	inproceedings	URL
BibTeX: @inproceedings{Xie2003, author = {L. Xie and S. Chang and A. Divakaran and H. Sun}, title = {Unsupervised discovery of multilevel statistical video structures using hierarchical hidden Markov models}, booktitle = {IEEE Intl. Conf. Multimedia and Expo (ICME)}, publisher = {IEEE Computer Society}, year = {2003}, pages = {29--32}, url = {citeseer.ist.psu.edu/xie03unsupervised.html} }
Yacoob, Y. & Davis, L.	Learned models for estimation of rigid and articulated human motion from stationary or moving camera [Abstract] [BibTeX]	2000	International Journal of Computer Vision	article
Abstract: We propose an approach for modeling, measurement and tracking of rigid and articulated motion as viewed from a stationary or moving camera. We first propose an approach for learning temporal-flow models from exemplar image sequences. The temporal-flow models are represented as a set of orthogonal temporal-flow bases that are learned using principal component analysis of instantaneous flow measurements. Spatial constraints on the temporal-flow are then incorporated to model the movement of regions of rigid or articulated objects. These spatio-temporal flow models are subsequently used as the basis for simultaneous measurement and tracking of brightness motion in image sequences. Then we address the problem of estimating composite independent object and camera image motions. We employ the spatio-temporal flow models learned through observing typical movements of the object from a stationary camera to decompose image motion into independent object and camera motions. The performance of the algorithms is demonstrated on several long image sequences of rigid and articulated bodies in motion
BibTeX: @article{Yacoob2000, author = {Yacoob, Yaser and Davis, Larry}, title = {Learned models for estimation of rigid and articulated human motion from stationary or moving camera}, journal = {International Journal of Computer Vision}, year = {2000}, volume = {36}, number = {1}, pages = {5--30} }
Zhang, D., Gatica-Perez, D., Bengio, S., McCowan, I. & Lathoud, G.	Modeling Individual and Group Actions in Meetings: A Two-Layer HMM Framework [BibTeX]	2004	cvprwInternational Conference on Computer Vision and Pattern Recognition	inproceedings	DOI
BibTeX: @inproceedings{Zhang2004, author = {Dong Zhang and Daniel Gatica-Perez and Samy Bengio and Iain McCowan and Guillaume Lathoud}, title = {Modeling Individual and Group Actions in Meetings: A Two-Layer HMM Framework}, booktitle = {International Conference on Computer Vision and Pattern Recognition}, journal = {cvprw}, publisher = {IEEE Computer Society}, year = {2004}, volume = {07}, pages = {117}, doi = {http://doi.ieeecomputersociety.org/10.1109/CVPR.2004.399} }
Zhang, W., Matsumoto, T., Liu, J., Chu, M. & Begole, B.	An intelligent fitting room using multi-camera perception. [BibTeX]	2008	Intelligent User Interfaces	inproceedings	URL
BibTeX: @inproceedings{Zhang2008, author = {Wei Zhang and Takashi Matsumoto and Juan Liu and Maurice Chu and Bo Begole}, title = {An intelligent fitting room using multi-camera perception.}, booktitle = {Intelligent User Interfaces}, publisher = {ACM}, year = {2008}, pages = {60-69}, url = {http://dblp.uni-trier.de/db/conf/iui/iui2008.html#ZhangMLCB08} }
Zhao, T.	Model-based Segmentation and Tracking of Multiple Humans in Complex Situations [BibTeX]	2003		phdthesis	URL
BibTeX: @phdthesis{Zhao2003a, author = {Zhao, Tao}, title = {Model-based Segmentation and Tracking of Multiple Humans in Complex Situations}, publisher = {University of South California}, year = {2003}, pages = {--}, url = {http://iris.usc.edu/~taozhao/research.html} }
Zhao, T. & Nevatia, R.	Tracking Multiple Humans in Crowded Environment [Abstract] [BibTeX]	2004	International Conference on Computer Vision and Pattern Recognition	inproceedings	DOI
Abstract: Tracking of humans in dynamic scenes has been an importanttopic of research. Most techniques, however, are limited to situationswhere humans appear isolated and occlusion is small.Typical methods rely on appearance models that must be acquiredwhen the humans enter the scene and are not occluded.We present a method that can track humans in crowded environments,with significant and persistent occlusion by makinguse of human shape models in addition to camera models, theassumption that humans walk on a plane and acquired appearancemodels. Experimental results and a quantitative evaluationare included.
BibTeX: @inproceedings{Zhao2004, author = {Zhao, Tao and Nevatia, Ram}, title = {Tracking Multiple Humans in Crowded Environment}, booktitle = {International Conference on Computer Vision and Pattern Recognition}, publisher = {IEEE Computer Society}, year = {2004}, volume = {02}, pages = {406-413}, doi = {http://doi.ieeecomputersociety.org/10.1109/CVPR.2004.253} }
Zhao, T. & Nevatia, R.	Bayesian Human Segmentation in Crowded Situations [Abstract] [BibTeX]	2003	International Conference on Computer Vision and Pattern Recognition	inproceedings
Abstract: Problem of segmenting individual humans in crowded situationsfrom stationary video camera sequences is exacerbatedby object inter-occlusion. We pose this problem as a"model-based segmentation" problem in which human shapemodels are used to interpret the foreground in a Bayesianframework. The solution is obtained by using an efficientMarkov chain Monte Carlo (MCMC) method which uses domainknowledge as proposal probabilities. Knowledge of variousaspects including human shape, human height, cameramodel, and image cues including human head candidates,foreground/background separation are integrated in one theoreticallysound framework. We show promising results andevaluations on some challenging data.
BibTeX: @inproceedings{Zhao2003, author = {Zhao, Tao and Nevatia, Ram}, title = {Bayesian Human Segmentation in Crowded Situations}, booktitle = {International Conference on Computer Vision and Pattern Recognition}, year = {2003}, pages = {--} }
Zhao, T. & Nevatia, R.	Stochastic Human Segmentation from a Static Camera [Abstract] [BibTeX]	2002	Workshop on Motion and Video Computing	inproceedings
Abstract: Segmenting individual humans in a high-density scene (e.g.,a crowd) acquired from a static camera is challenging mainlydue to object inter-occlusion (Fig.1). We define this problem asa "model-based segmentation" problem and the solution is obtainedusing a Markov chain Monte Carlo (MCMC) approach.Knowledge of various aspects including human shape, humanheight, camera model, and image cues including human headcandidates, foreground/background separation are integratedin a Bayesian framework. We show promising results on somechallenging data.
BibTeX: @inproceedings{Zhao2002, author = {Zhao, Tao and Nevatia, Ram}, title = {Stochastic Human Segmentation from a Static Camera}, booktitle = {Workshop on Motion and Video Computing}, year = {2002}, pages = {--} }
Zhao, T., Nevatia, R. & Lv, F.	Segmentation and Tracking of Multiple Humans in Complex Situations [Abstract] [BibTeX]	2001	International Conference on Computer Vision and Pattern Recognition	inproceedings
Abstract: Segmenting and tracking multiple humans is a challengingproblem in complex situations in which extended occlusion,shadow and/or reflection exists. We tackle this problemwith a 3D model-based approach. Our method includestwo stages, segmentation (detection) and tracking. Humanhypotheses are generated by shape analysis of the foregroundblobs using human shape model. The segmented human hypothesesare tracked with a Kalman filter with explicit handlingof occlusion. Hypotheses are verified while they aretracked for the first second or so. The verification is doneby walking recognition using an articulated human walkingmodel. We propose a new method to recognize walking usingmotion template and temporal integration. Experimentsshow that our approach works robustly in very challengingsequences.
BibTeX: @inproceedings{Zhao2001, author = {Zhao, Tao and Nevatia, Ram and Lv, Fengjun}, title = {Segmentation and Tracking of Multiple Humans in Complex Situations}, booktitle = {International Conference on Computer Vision and Pattern Recognition}, year = {2001}, pages = {--} }
	[BibTeX]			misc	URL
BibTeX: @misc{Ladybug,, url = {http://www.ptgrey.com/products/ladybug2/index.asp} }
	An Efficient Earth Mover's Distance Algorithm for Robust Histogram Comparison [BibTeX]	2007	IEEE Trans. Pattern Anal. Mach. Intell.	article	DOI
BibTeX: @article{1263456,, title = {An Efficient Earth Mover's Distance Algorithm for Robust Histogram Comparison}, journal = {IEEE Trans. Pattern Anal. Mach. Intell.}, publisher = {IEEE Computer Society}, year = {2007}, volume = {29}, number = {5}, pages = {840--853}, note = {Student Member-Haibin Ling and Member-Kazunori Okada}, doi = {http://dx.doi.org/10.1109/TPAMI.2007.1058} }
	MPEG-4 White Paper [BibTeX]	2005		unpublished
BibTeX: @unpublished{2005,, title = {MPEG-4 White Paper}, year = {2005} }

Created by JabRef on 11/08/2008.