Research map / 3D vision / 3D generation
PrimDiffusion: Volumetric Primitives Diffusion for 3D Human Generation
Zhaoxi Chen, Fangzhou Hong, Haiyi Mei, Guangcong Wang, Lei Yang, Ziwei Liu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Generate humans with volumetric primitives and diffusion rather than the compared GAN-based systems | PrimDiffusion · RenderPeople | FID_CLIP: 14.58 -> 12.11 | Improved |
| Generate humans with volumetric primitives and diffusion rather than the compared GAN-based systems | PrimDiffusion · RenderPeople | FPS: 22.97 -> 88.24 | Improved |
| Replace volumetric primitives with 64×64×64 voxel grids and 3D convolutions (ablation) | PrimDiffusion · RenderPeople | FID_CLIP: 12.11 -> 19.53 | Hurt |
| Replace primitive payloads with latent codes decoded by sparse convolutions (ablation) | PrimDiffusion · RenderPeople | FID: 17.95 -> 37.70 | Hurt |
| Use one uniformly shared delta scale factor instead of spatially varying delta scales (ablation) | Volumetric primitive fitting · Training identities, views, and poses | PSNR: 32.15 -> 18.35 | Hurt |
| Use template-mesh rotations and translations instead of identity-specific R,T (ablation) | Volumetric primitive fitting · Training identities, views, and poses | PSNR: 32.15 -> 31.70 | Hurt |
| Remove cross-modal attention (ablation) | Volumetric primitive fitting · Training identities, views, and poses | PSNR: 32.15 -> 28.39 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- EVA3D — EVA3D: Compositional 3D Human Generation from 2D Image Collections (ICLR 2023) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Human-3Diffusion: Realistic Avatar Creation via Explicit 3D Consistent Diffusion Models (NeurIPS 2024 · similarity 0.90)
- Morphable Diffusion: 3D-Consistent Diffusion for Single-image Avatar Creation (CVPR 2024 · similarity 0.89)
- HumanNorm: Learning Normal Diffusion Model for High-quality and Realistic 3D Human Generation (CVPR 2024 · similarity 0.88)
- PSHuman: Photorealistic Single-image 3D Human Reconstruction using Cross-Scale Multiview Diffusion and Explicit Remeshing (CVPR 2025 · similarity 0.88)
- En3D: An Enhanced Generative Model for Sculpting 3D Humans from 2D Synthetic Data (CVPR 2024 · similarity 0.87)
- SiTH: Single-view Textured Human Reconstruction with Image-Conditioned Diffusion (CVPR 2024 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
