Google Research Adj. Research Prof., Robotics Institute, CMU Email: rahuls@cs.cmu.edu http://www.cs.cmu.edu/~rahuls Research Interests - Computer vision - Machine learning		Rahul Sukthankar Research: Research bio Online publications Recent projects: CouchPotato -- Learning objects and actions from weakly supervised video. Paper: G. Hartmann et al., Weakly Supervised Learning of Object Segmentations from Web-Scale Video, ECCV Workshop on Web-scale Vision and Social Media, 2012. Video D-Nets: an alternative to patch-based descriptors like SIFT. CVPR'12 Paper: von Hundelshausen and Sukthankar, D-Nets: Beyond Patch-Based Image Descriptors Source code (integrates with OpenCV). Generalized Boundaries: ECCV'12 Paper: Leordeanu et al., Efficient Closed-Form Solution to Generalized Boundary Detection Source code SLIPstream -- scalable, low-latency perception Event Detection in Video Diamond -- Interactive exploration of non-indexed data Projector-Camera Systems (aka Smart Projectors) Events I am organizing: Short Courses/Tutorials Chair for IEEE CVPR 2013 [proposal submission deadline Dec. 3, 2012] AAAI Spring Symposium on Weakly Supervised Learning in Multimedia. [abstract submission deadline Oct 5, 2012] Special issue on Large-Scale Multimedia Corpus, Retrieval and Mining for IEEE Multimedia. Special issue on Labeling the World for IEEE Pervasive Computing Workshops Chair for IEEE CVPR 2009. Workshop on Analysis and Evaluation of Large-Scale Multimedia Collections, collocated with ICPR 2010. Workshop on Large-Scale Multimedia Mining and Retrieval, collocated with ACM Multimedia 2009. Workshop on Computer Vision for Humanoid Robots in Real Environments, collocated with ICCV 2009. International Workshop on Projector-Camera Systems (PROCAMS), collocated with ICCV, CVPR and SIGGRAPH. Workshop on cross-validation, bootstrap and model selection, collocated with NIPS 2000. Current student collaborators: Yicong Tian: Spatio-temporal DPMs (CVPR'13) Kevin Tang: CouchPotato - discriminative segment annotation (CVPR'13) Subh Bhattacharya: action recognition, photo recomposition Enrique Ortiz: co-training Former student collaborators: David Tsai: CouchPotato Matthias Grundmann: CouchPotato Pyry Matikainen: model recommendation for action recognition Yan Ke: PCA-SIFT; Music Identification; event detection in video Marius Leordeanu: learning shape for object category recognition Kevin Tang: (NSF REU student) object recognition Joel Jurik: (NSF REU student) object recognition David Todd: (CMU IFYRE student) information retrieval Dhruv Batra: class-specific affinities for image labeling Ming-yu (Robert) Chen: action recognition in video Liu Yang: Distance metric learning, medical imaging Derek Hoiem: Object-based Image Retrieval; Sound Object Detection Devi Parikh: Feature-based 3D reconstruction Stano Funiak: simultaneous localization and tracking Jonathan Huang Jay Summet: BurningWell -- localization using projected light S. Charles Brubaker: medical imaging Mark Ashdown and Matthew Flagg: Multi-planar Displays Han Chen: alignment of large-format projected displays Brandon Sanders: unsupervised object discovery Terence Sim: face recognition Trista Chen: image retrieval Earlier research (includes JKanji, face recognition and intelligent vehicles) Personal Interests: My wife, Gita, is an asst professor at UCF. Martial arts Games Scuba