Research map / Video / Video generation
Video4DGen: Enhancing Video and 4D Generation through Mutual Optimization
Yikai Wang, Guangce Liu, Xinzhou Wang, Zilong Chen, Jiafang Li, Xin Liang, Fuchun Sun, Jun Zhu
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 field initialization (ablation) | Video4DGen · SORA benchmark | PSNR: 15.42 -> 19.05 | Improved |
| add field initialization (ablation) | Video4DGen · Generated video Cat | PSNR: 20.15 -> 24.63 | Improved |
| add field initialization (ablation) | Video4DGen · SORA benchmark | MS-SSIM: 0.6167 -> 0.7323 | Improved |
| add dual branch refinement (ablation) | Video4DGen · SORA benchmark | PSNR: 18.57 -> 19.05 | Improved |
| add dual branch refinement (ablation) | Video4DGen · Generated video Cat | SSIM: 0.8196 -> 0.8432 | Improved |
| add our field initialization to Deformable-GS | Deformable-GS · Generated videos (avg of 30) | PSNR: 13.22 -> 22.63 | Improved |
| add our field initialization to SC-GS | SC-GS · Generated videos (avg of 30) | PSNR: 21.17 -> 24.75 | Improved |
| use full DGS method instead of SC-GS + field init | SC-GS + our field init. · Generated videos (avg of 30) | PSNR: 24.75 -> 27.30 | Improved |
| use sparse 25 control bones instead of 500 (ablation) | Video4DGen · generated videos | motion regularization | Improved |
| add warped-state geometric regularization (ablation) | Video4DGen · qualitative | structural integrity | Improved |
| add confidence map for 4D-guided video generation (ablation) | Video4DGen · 4D-guided generation | artifacts | Improved |
| use ground-truth-free pose via field initialization | Video4DGen · D-NeRF (without GT poses) | PSNR: 20.0486 -> 29.0624 | Improved |
Closest papers by meaning
- Vidu4D: Single Generated Video to High-Fidelity 4D Reconstruction with Dynamic Gaussian Surfels (NeurIPS 2024 · similarity 0.92)
- SV4D: Dynamic 3D Content Generation with Multi-Frame and Multi-View Consistency (ICLR 2025 · similarity 0.90)
- Robust Multi-Object 4D Generation for In-the-wild Videos (CVPR 2025 · similarity 0.90)
- Diffusion4D: Fast Spatial-temporal Consistent 4D generation via Video Diffusion Models (NeurIPS 2024 · similarity 0.89)
- 4Real: Towards Photorealistic 4D Scene Generation via Video Diffusion Models (NeurIPS 2024 · similarity 0.89)
- EG4D: Explicit Generation of 4D Object without Score Distillation (ICLR 2025 · 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.
