3D / Point Cloud
Objaverse
800K+ annotated 3D models for training 3D and multimodal models.
100K–1M·CC-BY·Point Cloud (LAS/PLY)
HeadsNet The basic concept is to train a FNN/MLP on vertex data of 3D heads so that it can then re-produce random 3D heads. This dataset uses the thispersondoesnotexist to triposr 6748 3D Heads dataset as a foundation. The heads dataset was collecting using the scraper Dataset Scraper.7z based on TripoSR which converts the 2D images from ThisPersonDoesNotExist into 3D meshes. (using this marching cubes improvement by thatname/zephyr) Vertex Normals need to be generated before we can… See the full description on the dataset page:
Source: Hugging Face Hub (tfnn/HeadsNet). Metadata imported from the dataset’s Hub tags.