1 paper
Zunlei Feng, Xinchao Wang, Yongming He +3
In this paper, we study a new representation-learning task, which we termed as disassembling object representations. Given an image featuring multiple objects, the goal of disassem…