Co-salient object detection with consensus mining and consistency cross-layer interactive decoding

Yanliang Ge, Jinghuai Pan, Junchao Ren, Min He, Hongbo Bi, Qiao Zhang

Abstract: Highlights•We propose a network for co-salient object detection, which combines consensus mining and consistency cross-layer interactive decoding.•The consensus cue mining module is designed to fully mine the cross-consensus clues among the co-salient objects in the image group.•The designed consistency cross-layer interactive decoder accepts features of different levels as input, so as to maintain the consistency of group consensus features.•A large number of experiments have demonstrated the effectiveness of our method.