1 paper · 1 filter
Tianhao Wu, Chuanxia Zheng, Frank Guan +2
Most image-based 3D object reconstructors assume that objects are fully visible, ignoring occlusions that commonly occur in real-world scenarios. In this paper, we introduce Amodal…