Papers of the day   All papers

Self-Supervised Scene De-occlusion


Goku Mohandas: ๐Ÿคฏ This recent work by @ccloy et al. blew my mind: Self-Supervised Scene De-occlusion (CVPR 2020) - Check out media/questions that Xiaohang (first author) has been answering on @madewithml:

10 replies, 1077 likes

Ziwei Liu: Our #CVPR2020 **oral** paper "Self-Supervised Scene De-occlusion": Paper: Code: Demo: - A self-supervised framework that can recover occluded objects & their spatial orders from a single RGB image.

0 replies, 219 likes

Chen Change Loy: Thanks @GokuMohandas for sharing our work! We already released the code at

0 replies, 99 likes

Insane: Self-Supervised Scene De-occlusion (CVPR 2020) Read More: Cc: @jblefevre60 @gp_pulipaka @fchollet @chipro @sallyeaves @FrRonconi @ingliguori @daviddegroof1 @PatrickGunz_CH @hwingo @ipfconline1 @TechNative @_TechMode @kcgcse @Ronald_vanLoon @grattonboy

1 replies, 15 likes

๐•Š๐•–๐•ฃ๐•˜ ๐•„๐•’๐•ครญ๐•ค ๐Ÿงข๐Ÿ“ˆ: #MachineLearning Wizardry! ๐Ÿง™โ€โ™‚๏ธ #AI #PyTorch

1 replies, 2 likes

akira: PCNets for self-supervised learning that complements the hidden parts of objects by adding individual order. By covering the image with other objects taken at random, the system learns to generate masks and complete the hidden parts.

0 replies, 1 likes


Found on Apr 07 2020 at

PDF content of a computer science paper: Self-Supervised Scene De-occlusion