Research map / 3D vision / 3D generation
FlashWorld: High-quality 3D Scene Generation within Seconds
Xinyang Li, Tengfei Wang, Zixiao Gu, Shengchuan Zhang, Chunchao Guo, Liujuan Cao
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 dual-mode cross-mode distillation (full model) vs MV-oriented diffusion only (ablation) | FlashWorld · T3Bench-200 text-to-3D | Q-Align IQA: 3.11 -> 4.12 | Improved |
| add cross-mode distillation vs 3D-oriented diffusion only (ablation) | FlashWorld · T3Bench-200 text-to-3D | Q-Align IQA: 2.61 -> 4.12 | Improved |
| use cross-mode distillation instead of MV-oriented distillation (ablation) | FlashWorld · T3Bench-200 text-to-3D | Q-Align IQA: 3.46 -> 4.12 | Improved |
| add cross-mode consistency (CMC) loss (ablation) | FlashWorld · T3Bench-200/DL3DV-200/WorldScore-200 | Q-Align IQA, CLIP metrics: 4.12 -> 4.12 (T3Bench Q-Align IQA); 4.02 -> 3.96 (DL3DV) | No clear effect |
| add out-of-distribution data co-training (ablation) | FlashWorld · T3Bench-200 text-to-3D | Q-Align IQA: 3.98 -> 4.12 | Improved |
| add out-of-distribution data co-training (ablation) | FlashWorld · T3Bench-200 text-to-3D | CLIP Aesthetic and Q-Align IAA: Q-Align IAA 2.50 (w/o OOD) -> 2.26 (full) | Hurt |
| FlashWorld vs Director3D on text-to-3D | Director3D · DL3DV-200 | Q-Align IQA: 2.51 -> 3.96 | Improved |
| FlashWorld vs Director3D on CLIP Score | Director3D · T3Bench-200 | CLIP Score: 27.84 -> 27.68 | No clear effect |
| FlashWorld vs Prometheus on CLIP Aesthetic and WorldScore-200 Q-Align IAA | Prometheus · T3Bench-200 CLIP Aesthetic | CLIP Aesthetic: 4.76 -> 4.49 | Hurt |
| FlashWorld vs LucidDreamer on WorldScore 3D Consistency | LucidDreamer · WorldScore static | 3D Consistency: 90.37 -> 85.87 | Hurt |
| FlashWorld vs WonderWorld on WorldScore Average | WonderWorld · WorldScore static | Average: 66.43 -> 68.72 | Improved |
| FlashWorld vs WonderWorld on Style Consistency | WonderWorld · WorldScore static | Style Consistency: 75.92 -> 81.52 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- Director3D — Director3D: Real-world Camera Trajectory and 3D Scene Generation from Text (NeurIPS 2024) · better on 5 of 7 shared results
- LucidDreamer — LucidDreamer: Towards High-Fidelity Text-to-3D Generation via Interval Score Matching (CVPR 2024) · 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
- Instant3D: Fast Text-to-3D with Sparse-view Generation and Large Reconstruction Model (ICLR 2024 · similarity 0.89)
- One-2-3-45++: Fast Single Image to 3D Objects with Consistent Multi-View Generation and 3D Diffusion (CVPR 2024 · similarity 0.89)
- A 3D Scene is Worth 1K Tokens: 3D-Grounded Representation for Scene Generation at Scale (NeurIPS 2026 · similarity 0.89)
- WonderWorld: Interactive 3D Scene Generation from a Single Image (CVPR 2025 · similarity 0.89)
- Director3D: Real-world Camera Trajectory and 3D Scene Generation from Text (NeurIPS 2024 · similarity 0.89)
- ViewDiff: 3D-Consistent Image Generation with Text-to-Image Models (CVPR 2024 · similarity 0.89)
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.
