Yifan Wang, Linjie Luo, Xiaohui Shen, Xing Mei: DynOcc: Learning Single-View Depth from Dynamic Occlusion Cues. 3DV 2020: 514-523