Research map / 3D vision / 3D generation
REPARO: Compositional 3D Assets Generation with Differentiable 3D Layout Alignment
Haonan Han, Rui Yang, Huan Liao, Jiankai Xing, Zunnan Xu, Xiaoming Yu, Junwei Zha, Xiu Li, Wanhua Li
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 |
|---|---|---|---|
| add optimal transport to RGB appearance loss (ablation) | REPARO (DreamGaussian) · GSO multi-object scenes (20) | CLIP score: 0.800 -> 0.828 | Improved |
| add optimal transport to RGB appearance loss (ablation) | REPARO (TripoSR) · GSO multi-object scenes (20) | CLIP score: 0.774 -> 0.819 | Improved |
| add depth to RGBXY OT appearance loss (ablation) | REPARO (DreamGaussian) · GSO multi-object scenes (20) | CLIP score: 0.828 -> 0.833 | Improved |
| add position and depth to RGB OT loss (ablation) | REPARO (TripoSR) · GSO multi-object scenes (20) | CLIP score: 0.819 -> 0.822 | No clear effect |
| add semantic loss L_s to appearance loss L_a (ablation) | REPARO (DreamGaussian) · GSO multi-object scenes (20) | LPIPS: 0.233 -> 0.234 | No clear effect |
| use only semantic loss without appearance loss (ablation) | REPARO (DreamGaussian) · GSO multi-object scenes (20) | CLIP score: 0.833 -> 0.809 | Hurt |
| add scale optimization to translation (ablation) | REPARO (DreamGaussian) · GSO multi-object scenes (20) | CLIP score: 0.832 -> 0.833 | No clear effect |
| add rotation optimization to translation and scale (ablation) | REPARO (DreamGaussian) · GSO multi-object scenes (20) | CLIP score: 0.833 -> 0.831 | Hurt |
| add rotation optimization to translation (ablation) | REPARO (DreamGaussian) · GSO multi-object scenes (20) | LPIPS: 0.232 -> 0.243 | Hurt |
| add rotation to translation and scale (ablation) | REPARO (TripoSR) · GSO multi-object scenes (20) | CLIP score: 0.822 -> 0.826 | Mixed |
| use TripoSR instead of DreamGaussian as 3D generator | REPARO · GSO multi-object scenes (20) | LPIPS: 0.234 -> 0.216 | Improved |
| use ground-truth assets instead of generated | REPARO · GSO multi-object scenes (20) | CLIP score: 0.822 -> 0.906 | Improved |
Later papers that compared with this method
1 later accepted paper reports REPARO as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- MIDI: Multi-Instance Diffusion for Single Image to 3D Scene Generation (CVPR 2025 · better on 4 of 4 shared results)
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.
- DreamGaussian — DreamGaussian: Generative Gaussian Splatting for Efficient 3D Content Creation (ICLR 2024) · better on 5 of 5 shared results
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
- PartCrafter: Structured 3D Mesh Generation via Compositional Latent Diffusion Transformers (NeurIPS 2025 · similarity 0.86)
- PartGen: Part-level 3D Generation and Reconstruction with Multi-view Diffusion Models (CVPR 2025 · similarity 0.86)
- Structured 3D Latents for Scalable and Versatile 3D Generation (CVPR 2025 · similarity 0.86)
- MeshAnything: Artist-Created Mesh Generation with Autoregressive Transformers (ICLR 2025 · similarity 0.86)
- Meta 3D AssetGen: Text-to-Mesh Generation with High-Quality Geometry, Texture, and PBR Materials (NeurIPS 2024 · similarity 0.85)
- One-2-3-45++: Fast Single Image to 3D Objects with Consistent Multi-View Generation and 3D Diffusion (CVPR 2024 · similarity 0.85)
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.
