Research map / Video / Video generation
Orthogonal Spatial-temporal Distributional Transfer for 4D Generation
Wei Liu, Shengqiong Wu, Bobo Li, Haoyu Zhao, Hao Fei, Mong-Li Lee, Wynne Hsu
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 spatial-temporal disentangling of latents (ablation) | Orster 4D model · Consistent4D (image-to-4D) | FVD: 596.7 -> 477.7 | Improved |
| add spatial channel (ablation) | Orster 4D model · Consistent4D (image-to-4D) | FVD: 501.1 -> 477.7 | Improved |
| add temporal channel (ablation) | Orster 4D model · Consistent4D (image-to-4D) | FVD: 549.3 -> 477.7 | Improved |
| add transferred temporal features in ST-HexPlane (ablation) | Orster 4D model · Consistent4D (image-to-4D) | CLIP-O: 0.81 -> 0.82 | No clear effect |
| add Orster distillation (ablation) | Orster 4D model · Consistent4D (image-to-4D) | CLIP-F: 0.40 -> 0.93 | Improved |
| add Orster distillation (ablation) | Orster 4D model · Consistent4D (image-to-4D) | FVD: 668.3 -> 477.7 | Improved |
| add preliminary 4D training (ablation) | Orster 4D model · Consistent4D (image-to-4D) | FVD: 601.6 -> 477.7 | Improved |
| add joint spatiotemporal Gaussian kernel (ablation) | Orster 4D model · Consistent4D (image-to-4D) | FVD: 566.4 -> 477.7 | Improved |
| add spatial-temporal consistency learning (ablation) | Orster 4D model · Consistent4D (image-to-4D) | CLIP-O: 0.72 -> 0.82 | Improved |
| add 4D construction training (ablation) | Orster 4D model · Consistent4D (image-to-4D) | FVD: 587.4 -> 477.7 | Improved |
| use spatial-only loss weight (lambda_o=1) vs both (ablation) | Orster 4D model · Consistent4D (image-to-4D) | CLIP-F: 0.59 -> 0.93 | 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.
- Diffusion4D — Diffusion4D: Fast Spatial-temporal Consistent 4D generation via Video Diffusion Models (NeurIPS 2024) · better on 6 of 6 shared results
- STAG4D — STAG4D: Spatial-Temporal Anchored Generative 4D Gaussians (ECCV 2024) · 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
- Diffusion4D: Fast Spatial-temporal Consistent 4D generation via Video Diffusion Models (NeurIPS 2024 · similarity 0.92)
- 4Diffusion: Multi-view Video Diffusion Model for 4D Generation (NeurIPS 2024 · similarity 0.92)
- SV4D: Dynamic 3D Content Generation with Multi-Frame and Multi-View Consistency (ICLR 2025 · similarity 0.89)
- Fused View-Time Attention and Feedforward Reconstruction for 4D Scene Generation (NeurIPS 2025 · similarity 0.89)
- Diffusion$^2$: Dynamic 3D Content Generation via Score Composition of Video and Multi-view Diffusion Models (ICLR 2025 · similarity 0.89)
- CAT4D: Create Anything in 4D with Multi-View Video Diffusion Models (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.
