Better 3D reconstruction without exact model match by using database of patches

Better 3D reconstruction without exact model match by using database of patches

Patch2CAD: Patchwise Embedding Learning for In-the-Wild Shape Retrieval from a Single Image
arXiv paper abstract https://arxiv.org/abs/2108.09368
arXiv PDF paper https://arxiv.org/pdf/2108.09368.pdf

3D perception of object shapes from RGB image input is fundamental towards semantic scene understanding