fcsod
8fc687aa152e8199fe9e73304d407bca-AuthorFeedback.pdf
Below, we conduct a point-point response to each comment. We will discuss them in the revised manuscript. ECCV18] apply the SPL processes to refine the saliency maps obtained in co-saliency detection. As we mentioned in the paper, these works share a similar spirit with our work. As for the difference, Y an et al. [23] define their problem on sparsely annotated video frames. It requires to generate image data for different problem domains.