David J. Crandall
School of Informatics and Computing
227 Informatics West, 901 E Tenth St
Bloomington, IN 47408
|I work in computer vision, the area of computer
science concerned with automatically inferring semantic meaning from
images -- teaching computers to "see." More generally, I am interested
in problems that involve analyzing and modeling large amounts of
uncertain data, like mining data from social networking websites.
I joined the School of Informatics and
Computing at Indiana University in Fall 2010. I'm on the faculty
of the Computer Science, Informatics, and Cognitive Science programs.
As a graduate student and postdoc in Computer Science at Cornell University, I worked with Professor
Dan Huttenlocher on
statistical part-based object recognition algorithms. I also worked
with Professor Jon
Kleinberg on modeling and mining data from online social
networks. I completed the Ph.D. degree in August 2008.
Before Cornell, I spent two years in the research labs
of Eastman Kodak Company. There I
worked mostly on image understanding and enhancement algorithms
for both consumer and medical images.
As an undergraduate and Masters student
at Penn State University, I worked
with Professor Rangachar
Kasturi on content-based video indexing. We worked mostly on
detecting, tracking and recognizing text in video (both captions and
text appearing naturally in a scene).
This is a partial list of publications. Please see my lab's publications page or my Google Scholar page for a more complete listing.
- – "SfM with MRFs: Discrete-Continuous Optimization for Large-Scale Structure from Motion," in PAMI 2013 (with A. Owens, N. Snavely, D. Huttenlocher)
- – "Discrete-Continuous Optimization for Large-Scale Structure from Motion," in CVPR 2011 (with A. Owens, N. Snavely, D. Huttenlocher)
Runner-up best paper!
First-person and opportunistic imagery
- – "PlaceAvoider: Steering First-Person Cameras away from Sensitive Spaces," in NDSS 2014 (with R. Templeman, M. Korayem, A. Kapadia)
- – "PlaceRaider: Virtual Theft in Physical Spaces with Smartphones," in NDSS 2013 (with R. Templeman, Z. Rahman, A. Kapadia)
- – "Discovering localized attributes for fine-grained recognition," in CVPR 2012 (with K. Duan, D. Parikh, K. Grauman)
- – "A Multi-layer Composite Model for Human Pose Estimation," in BMVC 2012 (with K. Duan, D. Batra)
- – "Landmark classification in large-scale image collections," in ICCV 2009 (with Y. Li and D. Huttenlocher)
- – "Composite models of objects and scenes for category recognition," in CVPR 2007 (with D. Huttenlocher)
- – "Weakly-supervised learning of part-based spatial models for visual object recognition," in ECCV 2006 (with D. Huttenlocher)
- – "Robust Color Object Detection using Spatial-Color
Joint Probability Functions," in IEEE
Transactions on Image Processing, 2006 (with J. Luo)
[CVPR 2004 version]
- – "Part-based statistical models for visual object class recognition," 2008 (Ph.D. thesis)
Mining and modeling online social networks
- – "De-anonymizing users across heterogeneous social computing platforms," in ICWSM 2013 (with M. Korayem)
- – "Mining photo-sharing websites to study ecological phenomena," in WWW 2012 (with H. Zhang, M. Korayem, G. LeBuhn)
- – "Beyond co-occurrence: Discovering and visualizing tag relationships from geo-spatial and temporal similarities," in WSDM 2012 (with H. Zhang, M. Korayem, E. You)
- – "Inferring Social Ties from Geographic Coincidences," in Proc. National Academy of Sciences (PNAS), 8 December 2010 (with L. Backstrom, D. Cosley, S. Suri, D. Huttenlocher, J. Kleinberg)
- – "Feedback Effects between Similarity and Social Influence in Online Communities," in KDD 2008 (with D. Cosley, J. Kleinberg, D. Huttenlocher, S. Suri)
- – "Understanding embodied visual attention in child-parent interaction," in ICDL 2013 (with S. Bambach, C. Yu)
- – "Psychophysical study of image orientation perception,"
in Spatial Vision, 2003, pp. 429-457. (with J. Luo,
A. Singhal, M. Boutell and R. Gray.)
Content-based video indexing
- – "Extraction of special effects caption text events from
digital video," in International Journal on Document Analysis and
Recognition, 2002 (with S. Antani and R. Kasturi)
[M.S. thesis version]
I'm teaching two courses in Fall 2013:
- – Info I427, Search Informatics. Students learn about information retrieval and web search, and build their own search engines as the course project.
- – Info I399, Research Methods for Undergraduates. Students study basic research methods for informatics and computing fields, and conduct a semester-long research project.
- – Taught CS B553, Probabilistic Approaches to Artificial Intellignece (Spring 2013)
- – Taught I427, Search Informatics (Fall 2011, Spring 2011, Fall 2012)
- – Taught CS B657, Computer Vision, (Spring 2012, Fall 2010)
- – Taught INFO 2950, Mathematical Methods for Information Science, Fall 2008 (at Cornell)
- – Taught CS 113, Introduction to C, Fall 2007 and Spring 2008 (at Cornell)
- – Taught CS 211, Algorithms and Data Structures in Java, Summer 2007 (at Cornell)
- – T.A. for CS 664, Computer Vision, Fall 2003 (at Cornell)
- – Taught CSE 275, Digital Design Lab, Fall 2000 and Spring 2001 (at Penn State)
These are other projects I've worked on that were never polished enough
to be publishable, but that someone somewhere might still find interesting. :-)
| During the summer, my main obsession is cycling. I
ride for fun only, and I have no delusions that I'm actually any
good. Recent trips include Ithaca to Lake Erie (250 miles), Ithaca to
State College (200 miles), and around Cayuga Lake (85 miles). I like
to ski during the winter (aka the majority of the year in upstate New
York), although I'm even worse at that.
I'm on a perpetual quest to become fluent in Spanish. In addition to
taking classes in high school and throughout college and graduate
school, I've studied abroad in Spain and Mexico, and I've also traveled
to Puerto Rico and El Salvador. Despite all that effort, my Spanish
is still remarkably lousy. :)
For the last few years I have volunteered as an income tax return
at Alternatives Federal Credit
Union. Believe it or not, it's surprisingly fun.