1 paper
Xiaoyun Zheng, Liwei Liao, Jianbo Jiao +2
Self-supervised Object Segmentation (SOS) aims to segment objects without any annotations. Under conditions of multi-camera inputs, the structural, textural and geometrical consist…