上一条: Visual tracking via robust multi-task multi-feature joint sparse representation
下一条: A unified framework for interactive image segmentation via Fisher rules