Research map / 3D vision / 3D generation
OneWorld: Taming Scene Generation with 3D Unified Representation Autoencoder
Sensen Gao, Zhaoqing Wang, Qihang Cao, Dongdong Yu, Changhu Wang, Tongliang Liu, Mingming Gong, Jiawang Bian
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 cross-view correspondence (CVC) loss (ablation) | OneWorld · RealEstate10K (1-view NVS) | PSNR: 19.10 -> 21.57 | Improved |
| add manifold-drift forcing (ablation) | OneWorld · RealEstate10K (1-view NVS) | PSNR: 20.59 -> 21.57 | Improved |
| add appearance injection branch (ablation) | 3D-URAE · 3D-URAE reconstruction | PSNR: 21.14 -> 28.19 | Improved |
| add semantic distillation (ablation) | OneWorld · Generation (semantic distillation ablation) | PSNR: 17.45 -> 21.57 | Improved |
| raise CVC threshold tau from 0.80 to 0.90 (ablation) | OneWorld CVC · RealEstate10K (1-view NVS) (CVC hparam, 10K steps) | PSNR: 17.62 -> 19.10 | Improved |
| raise CVC threshold tau from 0.90 to 0.95 (ablation) | OneWorld CVC · RealEstate10K (1-view NVS) (CVC hparam, 10K steps) | PSNR: 19.10 -> 18.21 | Hurt |
| raise CVC weight from 0.20 to 0.40 (ablation) | OneWorld CVC · RealEstate10K (1-view NVS) (CVC hparam, 10K steps) | PSNR: 19.10 -> 17.41 | Hurt |
| raise semantic weight lambda_sem from 0.10 to 0.20 (ablation) | 3D-URAE · 3D-URAE reconstruction | PSNR: 28.19 -> 26.58 | Hurt |
| increase distillation margin from 0.00 to 0.05 (ablation) | 3D-URAE · 3D-URAE reconstruction | PSNR: 26.67 -> 28.19 | Improved |
| use OneWorld over FlashWorld | OneWorld · DL3DV-10K (1-view NVS) | PSNR: 16.02 -> 17.19 | Improved |
| use x0-prediction instead of v-prediction in 3D token space (ablation) | OneWorld DiT · single-scene overfit | fitting speed | Improved |
| use 3D-URAE for 8-view feed-forward reconstruction vs DepthSplat | 3D-URAE · DL3DV (8-view feed-forward 3DGS) | LPIPS: PSNR 23.39 -> 24.68; LPIPS 0.135 (DepthSplat) vs 0.140 | Mixed |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- Gen3R — Gen3R: 3D Scene Generation Meets Feed-Forward Reconstruction (CVPR 2026) · better on 4 of 4 shared results
- Gen3C — GEN3C: 3D-Informed World-Consistent Video Generation with Precise Camera Control (CVPR 2025) · better on 4 of 4 shared results
- DepthSplat — DepthSplat: Connecting Gaussian Splatting and Depth (CVPR 2025) · better on 2 of 2 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
- A 3D Scene is Worth 1K Tokens: 3D-Grounded Representation for Scene Generation at Scale (NeurIPS 2026 · similarity 0.90)
- LT3SD: Latent Trees for 3D Scene Diffusion (CVPR 2025 · similarity 0.89)
- HarmonyView: Harmonizing Consistency and Diversity in One-Image-to-3D (CVPR 2024 · similarity 0.88)
- PartCrafter: Structured 3D Mesh Generation via Compositional Latent Diffusion Transformers (NeurIPS 2025 · similarity 0.88)
- One-2-3-45++: Fast Single Image to 3D Objects with Consistent Multi-View Generation and 3D Diffusion (CVPR 2024 · similarity 0.88)
- Ouroboros3D: Image-to-3D Generation via 3D-aware Recursive Diffusion (CVPR 2025 · similarity 0.88)
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.
