Research Overview

My research interests mainly lie in the areas of computer vision and machine learning. I am interested in developing statistical models that can accurately describe the imaging process but still allow for efficient inference algorithm to solve problems of visual inference.

My current research focuses on optical flow estimation. Enabled by naturalistic sequences associated with ground truth flow fields, we have learned a complete probabilistic model of optical flow (ECCV 2008 paper). We have also implemented the most widely cited Horn & Schunck method using currently common practices in optical flow estimation and found it performs on par with some more sophisticated methods published in the last few years (Middlebury flow evaluation results).

 

Publications

Journal papers

Deqing Sun and Wai-Kuen Cham. “Postprocessing of Low Bit Rate Block DCT Coded Images based on a Fields of Experts Prior.” IEEE Trans. Image Proc., 16(11), pp. 2743- 2751, Nov. 2007. [pdf]

Deqing Sun, Stefan Roth, and Michael J. Black. "A Quantitative Analysis of Current Practices in Optical Flow Estimation and the Principles Behind Them". International Journal of Computer Vision (IJCV), 2013 [pdf] [Matlab code] [From publisher]

Ce Liu and Deqing Sun. "On Bayesian Adaptive Video Super Resolution". IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), to appear. [ pdf ]

Conference papers

Deqing Sun, Jonas Wulff, Erik B. Sudderth, Hanspeter Pfister, and Michael J. Black. "A Fully-Connected Layered Model of Foreground and Background Flow". IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2013. [pdf] [Sup. Mat.] [Code]

Deqing Sun and Ce Liu. "Non-causal Temporal Prior for Video Deblocking". European Conferenceon Computer Vision (ECCV), 2012. [pdf]

Deqing Sun, Erik B. Sudderth, and Michael J. Black. "Layered Segmentation and Optical Flow Estimation over Time". IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2012. [pdf] [Sup. Mat.] [Poster] [Data]

Ce Liu and Deqing Sun. "A Bayesian Approach to Adaptive Video Super Resolution". IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2011. Oral presentation.[pdf]

Deqing Sun, Erik B. Sudderth, and Michael J. Black. "Layered Image Motion with Explicit Occlusions, Temporal Consistency, and Depth Ordering". Neural Information Processing Systems (NIPS), 2010. [pdf] [Spotlight] [Poster]

Deqing Sun, Stefan Roth, and Michael J. Black. "Secrets of Optical Flow Estimation and Their Principles". In IEEE Conf. on Computer Vision and Pattern Recognition (CVPR). 2010 [pdf] [MATLAB code] [Results on Middlebury training set] [Spotlight] [Poster]

Deqing Sun, Stefan Roth, J.P. Lewis, and Michael J. Black. "Learning Optical Flow". In Proc. of the European Conference on Computer Vision (ECCV), 2008. [pdf] [Talk slides] [Extra training data]

Deqing Sun and Wai-Kuen Cham. “An Effective Postprocessing Method for Low Bit Rate Block DCT Coded Images.” IEEE Int. Conf. Acoustics, Speech, and Signal Proc. (ICASSP’07) 2007 [pdf].

Yifeng Jiang, Jun Xie, Deqing Sun, and Hung-Tat Tsui. “ Shape Registration by Simultaneously Optimizing Representation and Transformation”. Int. Conf. on Medical Image Computing and Computer Assisted Intervention (MICCAI), 2007.

Deqing Sun, Wai-Kuen Cham, and Junhao Xie. “A New Postprocessing Algorithm of Low Bit Rate JPEG Coded Images.” SPIE Int. Conf. on Visual Comm. and Image Proc. (VCIP) pages 1204 - 1214, Bei Jing, China, July 2005.

Thesis

Deqing Sun "From Pixels to Layers: Joint Motion Estimation and Segmentation". Ph.D. Dissertation, Brown University, May 2013. [pdf]

Deqing Sun. “Postprocessing of Images Coded Using Block DCT at Low Bit Rates.” M.Phil. thesis, The Chinese University of Hong Kong, July 2007.

Copyright Notice

All materials are presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors and other holders (publishers). These works may not be reposted without the explicit permission of the copyright holder.