I am an Associate Professor in the Department of Computer Science, Rutgers University

 

Areas of Interest

Research Focus: Computer Vision, Visual Learning, Human motion analysis, Activity Recognition, Tracking.

Wider interest: Multimedia Computing, Image and Video Processing, Image and Video Databases, Machine Learning, Neural Modeling and AI.

Research Statement

The focus of computer vision research is on recovering and understanding knowledge about the three-dimensional world from two-dimensional images. Such problems have proved to be extremely difficult. A central issue in computational vision is object and scene representation. Scientists have investigated different representations, varying from object-centered 3D geometric representations to viewer-centered representations. The problem is more challenging for dynamic objects. My research interest is on investigating representations of moving objects (articulated and deformable). In particular, I am interested in studying representations that explain the visual appearance of dynamic objects without the need for explicit 3D geometric models. My objective is to learn representations for the shape and the appearance of moving (dynamic) objects that support tasks such as synthesis, pose recovery, reconstruction and tracking. In particular, my research is focused on this problem within the human motion analysis context. Human motion analysis is a challenging problem with many potential applications such as visual surveillance, human-machine interface, video archival and retrieval, computer graphics animation, autonomous driving, virtual reality, etc. [Read more about my research…]

NEWS & RECENT ACTIVITIES

New Algorithm for Moving Camera Background Subtraction - “Online Moving Camera Background Subtraction” Published in ECCV2012 - Watch Demos Here

I am serving as an area chair and publication chair in the 10th IEEE Conference on Automatic Face and Gesture Recognition FG2013.

Check out our approach for satellite image based GPS-free vehicle localization - Published in ICRA 2012 - Watch Demos Here

Book Chapter on Background Subtraction (Tutorial and Survey)  “Figure-ground segmentation - pixel-based” in “Guide to Visual Analysis of Humans: Looking at People” To be published in 2011 by Springer

Two CVPR’11 papers: “Line-based relative pose estimation” with Ali Elqursh and “Supervised Hypergraph Labeling” with Toufiq Parag

Two CVPR’10 papers: “Putting Local Features on a Manifold” and “One-Shot Multi-Set Non-rigid Feature-Spatial Matching” with Marwan Torki

A short tutorial on skin detection for face detection - to appear in Encyclopedia of Biometrics by Springer.

Three new journal papers to appear in early 2009 in PAMI, IJCV, and CVIU        

       -  “Tracking People on a Torus” - TPAMI - march 09

Book Chapter “The Role of Manifold Learning in Human Motion Analysis” Human Motion - Understanding, Modeling, Capture and Animation. Springer – Computational Imaging Series