Visit ShanghaiTech University | 中文
HOME > News and Events > Seminars
Recent Work on Image Segmentation and Salient Object Detection
Date: 2016/6/3             Browse: 271

Recent Work on Image Segmentation and Salient Object Detection

Speaker: Yizhou Yu

Time: Jun 3, 3:00pm - 4:20pm.

Location: Room 105, H2 Center


I first present a new nonlinear embedding, called piecewise flat embedding, for image segmentation. Piecewise flat embedding attempts to identify segment boundaries while significantly suppressing variations within segments. We adopt an L1-regularized energy term in the formulation to promote sparse solutions. We further devise an effective two-stage numerical algorithm based on Bregman iterations to solve the proposed embedding. Piecewise flat embedding can be easily integrated into existing image segmentation frameworks. Experiments on BSDS500 indicate that segmentation algorithms incorporating this embedding can achieve significantly improved results.

Salient object detection has recently witnessed substantial progress due to powerful features extracted using deep convolutional neural networks (CNNs). However, existing CNN-based methods operate at the patch level instead of the pixel level. Resulting saliency maps are typically blurry, especially near the boundary of salient objects. Furthermore, image patches are treated as independent samples even when they are overlapping, giving rise to significant redundancy in computation and storage. In the second part of this talk, I present an end-to-end deep contrast network to overcome the aforementioned limitations. Our deep network consists of two complementary components, a pixel-level fully convolutional stream and a segment-wise spatial pooling stream. Finally, a fully connected CRF model can be optionally incorporated to improve spatial coherence and contour localization in the fused result from these two streams. Experimental results demonstrate that our deep network significantly improves the state of the art.

This is joint work with Chaowei Fang, Guanbin Li and Zicheng Liao..


Yizhou Yu received the PhD degree from the University of California at Berkeley in 2000. He is currently a professor in the Department of Computer Science, The University of Hong Kong, and an adjunct professor at the University of Illinois, Urbana-Champaign. He received the National Science Foundation CAREER Award, and the Best Paper Awards at 2005 and 2011 ACM SIGGRAPH/EG Symposium on Computer Animation. Prof Yu has served on the editorial board of IEEE Transactions on Visualization and Computer Graphics, Computer Graphics Forum, The Visual Computer and International Journal of Software and Informatics. He has also served on the program committee of many leading international conferences, including SIGGRAPH, SIGGRAPH Asia, and International Conference on Computer Vision. 

SIST-Seminar 16040