|Topic:||Label propagation for one-shot video object segmentation|
|Supervisor:||Georgios Tolias, Ph.D.|
|Announce as:||Diplomová práce, Semestrální projekt|
|Description:||One-shot video object segmentation aims to separate a target object from a video sequence, given the mask in a single frame. This project will explore a simple yet strong transductive method that relies on a proximity graph between local regions of frames at different times and performs label propagation. Features in the proximity graph are derived from deep convolutional neural networks. The goal of the project is reproduce results of the work of Zhang et al. CVPR 2020 and seek improvements in the propagation scheme and the feature quality.
|Bibliography:||Zhang, Yizhuo, Zhirong Wu, Houwen Peng, and Stephen Lin. "A Transductive Approach for Video Object Segmentation." In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 6949-6958. 2020.
Zhou, Dengyong, Olivier Bousquet, Thomas N. Lal, Jason Weston, and Bernhard Schölkopf. "Learning with local and global consistency." In Advances in neural information processing systems, pp. 321-328. 2004.