A source of multi-view orthographic renderings of a high-quality subset of the Objaverse collection, featuring 1024x1024 resolution images. Each entry includes four distinct modalities: RGB, Depth, Normal maps, and Camera parameters for 10 views per object.
Use Cases
- Train multi-view diffusion models using the RGB and Camera parameter data
- Develop depth estimation algorithms by comparing RGB inputs against the provided Depth maps
- Perform surface geometry analysis using the Normal map modality
- Fine-tune 3D generation adapters like MV-Adapter using the multi-view orthographic image sets
Strengths
- 1024x1024 resolution for all rendered image modalities
- Includes RGB, Depth, Normal, and Camera data for every view
- Utilizes orthographic projection views rather than perspective cameras
- Derived from a filtered high-quality subset of the Objaverse 3D model repository