I am interested in geometric problems in Computer Vision which involve multiple cameras inferring 3D structure from images.

My current research deals with 3D aspects of computer vision - namely, camera calibration, multi-view reconstruction, stereo, shape-from-silhouettes, application of discrete optimization in computer vision and implementation of vision algorithms on graphics hardware.

Current Research

• Piecewise Planar Stereo [Project page] New!
• Interactive Image-based Modeling for Architectural Scenes [Project page] New!
• Multi-view Stereo [Project page]
• Visual Hulls and Multi-view Reconstruction.
• Graphics Hardware implementation of low-level vision algorithms. [GPU_KLT Tracker Project page]  
• Self Calibration of Pan Tilt Zoom (PTZ) Cameras. [Video Summary]
• Camera Network Calibration and Synchronization using Dynamic Silhouettes [Code Available] New!
• Computing High Resolution Mosaics using PTZ cameras. [Project page]

Journal Papers

1. Sudipta N Sinha and Marc Pollefeys, "Pan-Tilt-Zoom Camera Calibration and high-resolution mosaic generation", Computer Vision and Image Understanding (CVIU), special issue on Omnidirectional Cameras and Camera Networks, Volume 103, Issue 3 (September 2006), pp. 170-183.


2. Sudipta N Sinha, Jan-Michael Frahm, Marc Pollefeys and Yakup Genc, "Feature Tracking and Matching in Video Using Programmable Graphics Hardware", (accepted, March 2007), Machine Vision and Applications (MVA), (available online - DOI 10.1007/s00138-007-0105-z)

Conference Papers

1. Sudipta N. Sinha, Drew Steedly and Richard Szeliski, "Piecewise Planar Stereo for Image-based Rendering", (to appear) ICCV 2009 at Kyoto, Sep 2008. [pdf] [webpage]


2. Sudipta N. Sinha, Drew Steedly, Richard Szeliski, Maneesh Agarwala and Marc Pollefeys, "Interactive 3D Architectural Modeling from Unordered Photo Collections", (to appear) SIGGRAPH ASIA 2008 at Singapore, Dec2008. [pdf] [webpage]


3. Sudipta N Sinha, Philippos Mordohai and Marc Pollefeys, "Multi-View Stereo via Graph Cuts on the Dual of an Adaptive Tetrahedral Mesh", (to appear) In Proc. of ICCV, 2007, at Rio de Janiero, Brazil. [pdf]


4. Li Guan, Sudipta N. Sinha, Jean-Sebastien Franco and Marc Pollefeys, "Visual Hull Construction in the Presence of Partial Occlusion", 3DPVT, Chapel Hill, June 2006. [pdf]


5. Sudipta N Sinha and Marc Pollefeys, "Multi-view Reconstruction using Photo-consistency and Exact Silhouette Constraints: A Maximum-Flow Formulation", In Proc. of ICCV, 2005, at Beijing, China. [pdf]


6. Sudipta N Sinha, Marc Pollefeys and Seon Joo Kim, "High-Resolution Multiscale Panoramic Mosaics from Pan-Tilt-Zoom Cameras", In Proc. of 4th Indian Conference on Computer Vision, Graphics and Image Processing, 2004, at Kolkata, India. [pdf]


7. Sudipta N Sinha and Marc Pollefeys, "Synchronization and Calibration of Camera Networks from Silhouettes", ICPR 2004, at Cambridge,UK. [pdf]


8. Sudipta N Sinha, Marc Pollefeys and Leonard McMillan, "Camera Network Calibration from Dynamic Silhouettes", CVPR 2004, at Washington DC. [pdf]


9. Marc Pollefeys and Sudipta Sinha, "Iso-disparity surfaces for general stereo configurations", ECCV 2004, at Prague. [pdf]


10. Kelshikar, N., X. Zabulis, K. Daniilidis, V. Sawant, S. Sinha, T. Sparks, S. Larsen, H. Towles, K. Mayer-Patel, H. Fuchs, J. Urbanic, K. Benninger, R. Reddy, and G. Huntoon. "Real-time Terascale Impementation of Tele-immersion," International Conference on Computation Science 2003. Melbourne, Australia. (June 2-4, 2003). [pdf]


11. Soumyadeep Paul, Sudipta N Sinha and Amitabha Mukerjee, Virtual Kathakali : Gesture Driven Metamorphosis, Proceedings of International Conference on Knowledge Based Computer Systems, Mumbai, India, Dec '98. [pdf]

Workshop Papers

1. Sudipta N Sinha, Jan-Michael Frahm, Marc Pollefeys and Yakup Genc, "GPU-Based Video Feature Tracking and Matching", EDGE 2006, workshop on Edge Computing Using New Commodity Architectures, Chapel Hill, May 2006. [pdf]


2. Sudipta N Sinha and Marc Pollefeys, "Towards Calibrating a Pan-Tilt-Zoom Camera Network", OMNIVIS 2004, workshop on Omnidirectional Vision and Camera Networks held in conjunction with ECCV 2004. [pdf]


3. Towles, Herman, Sang-Uok Kum, Travis Sparks, Sudipta Sinha, Scott Larsen and Nathan Beddes. "Transport and Rendering Challenges of Multi-Stream 3D Tele-Immersion Data," NSF Lake Tahoe Workshop on Collaborative Virtual Reality and Visualization (CVRV 2003) Tahoe City, CA, USA (October 26-28, 2003)." [pdf]

Invited Papers

1. A. Akbarzadeh, J.-M. Frahm, P. Mordohai, B. Clipp, C. Engels, D. Gallup, P. Merrell, M. Phelps, S.N. Sinha, B. Talton, L. Wang, Q. Yang, H. Stewenius, R. Yang, G. Welch, H. Towles, D. Nister and M. Pollefeys, "Towards Urban 3D Reconstruction From Video",3DPVT, Chapel Hill, June 2006. [pdf]


2. Sudipta N Sinha and Marc Pollefeys, "Visual-Hull Reconstruction from Uncalibrated and Unsynchronized Video Streams", 2nd International Symposium on 3D Data Processing, Visualization & Transmission, 2004, at Thessaloniki, Greece. [pdf]

Others

1. Sudipta Sinha, "Calibration of a heterogeneous network of color and depth cameras", one of the 10 selected entries in Phase I of the Canesta Design Contest, Dec 2004. [pdf]


2. M. Pollefeys, S.N. Sinha and J. Yan, "Calibration and Shape Recovery from Videos of Dynamic Scenes", (Invited Book Chapter) Computational Vision in Neural and Machine Systems, L. Harris and M. Jenkins (Eds.), Cambridge University Press, 2007.

Video Proceedings

1. Sudipta N Sinha and Marc Pollefeys, "Synchronization and Calibration of a Camera Network for 3D Event Reconstruction from Live Video", VPCVPR 2005, at San Diego. [pdf] (1 page)


2. SriRam Thirthala, Sudipta N Sinha and Marc Pollefeys, "Calibration of Pan-Tilt-Zoom (PTZ) Cameras and Omni-directional Cameras", VPCVPR 2005, at San Diego. [pdf] (1 page)

Integrative Paper

1. Sudipta N Sinha, "Graph Cut Algorithms in Vision, Graphics and Machine Learning", Integrative Paper, November, 2004, UNC Chapel Hill. [pdf]

Past Research

• Multiple PC Rendering Architecture for Realtime Tele-immersion viewer.  September 2002 - present
 
My focus has been to design and implement a distributed architecture for rendering around 1.5 million point cloud datasets in real-time from multiple camera streams. We have been able to interactively render stream of depth maps from different camera sets reconstructing a complete scene at an update rate of upto 5 fps. A stereo display system and a head tracking system provides the infrastructure to immerse the user in the rendered scene and give the illusion of  'Being there'.