Fetching the paper…

Self-supervised Single-view 3D Reconstruction via Semantic Consistency · Around