Research map / Video / Video generation
STAG4D: Spatial-Temporal Anchored Generative 4D Gaussians
Yifei Zeng, Yanqin Jiang, Siyu Zhu, Yuanxun Lu, Youtian Lin, Hao Zhu, Weiming Hu, Xun Cao, Yao Yao
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 anchor (Zero123++ multi-view input) to baseline (ablation) | STAG4D · Consistent4D video-to-4D | FVD: 1369.65 -> 1198.23 | Improved |
| add temporal anchor attention fusion on top of spatial anchor (ablation) | STAG4D · Consistent4D video-to-4D | FVD: 1198.23 -> 992.21 | Improved |
| add adaptive densification (top 2.5% gradient) (ablation) | STAG4D · Consistent4D video-to-4D | FVD: 1453.53 -> 992.21 | Improved |
| replace cross-frame attention with first-frame anchored spatial-temporal attention (ablation) | Zero123++ in STAG4D · Consistent4D video-to-4D | FID-VID: 63.32 -> 52.58 | Improved |
| use fixed densification threshold instead of adaptive (ablation) | STAG4D · Bird and Duck cases | qualitative | Hurt |
| use cross-frame attention instead of spatial-temporal anchored attention (ablation) | Zero123++ · Consistent4D video-to-4D | CLIP: 0.901 vs 0.899 (spatial only) | No clear effect |
Later papers that compared with this method
11 later accepted papers report STAG4D as a baseline in their results tables; 11 of them report a better number on most of the results they share with it.
- 4DSTR: Advancing Generative 4D Gaussians with Spatial-Temporal Rectification for High-Quality and Consistent 4D Generation (AAAI 2026 · better on 7 of 7 shared results)
- 4DVD: Cascaded Dense-view Video Diffusion Model for High-quality 4D Content Generation (IJCV 2026 · better on 4 of 4 shared results)
- Turbo4DGen: Ultra-Fast Acceleration for 4D Generation (ICML 2026 · better on 3 of 3 shared results)
- Orthogonal Spatial-temporal Distributional Transfer for 4D Generation (AAAI 2026 · better on 2 of 2 shared results)
- Not All Frame Features Are Equal: Video-to-4D Generation via Decoupling Dynamic-Static Features (ICCV 2025 · better on 10 of 10 shared results)
- SV4D: Dynamic 3D Content Generation with Multi-Frame and Multi-View Consistency (ICLR 2025 · better on 4 of 4 shared results)
- MVTokenFlow: High-quality 4D Content Generation using Multiview Token Flow (ICLR 2025 · better on 3 of 4 shared results)
- DIMO: Diverse 3D Motion Generation for Arbitrary Objects (ICCV 2025 · better on 3 of 3 shared results)
3 more not listed.
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.
- Consistent4D — Consistent4D: Consistent 360° Dynamic Object Generation from Monocular Video (ICLR 2024) · better on 3 of 3 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 0.91)
- 4Diffusion: Multi-view Video Diffusion Model for 4D Generation (NeurIPS 2024 · similarity 0.91)
- Align Your Gaussians: Text-to-4D with Dynamic 3D Gaussians and Composed Diffusion Models (CVPR 2024 · similarity 0.91)
- Generative Sparse-View Gaussian Splatting (CVPR 2025 · similarity 0.91)
- Diffusion$^2$: Dynamic 3D Content Generation via Score Composition of Video and Multi-view Diffusion Models (ICLR 2025 · similarity 0.90)
- Fused View-Time Attention and Feedforward Reconstruction for 4D Scene Generation (NeurIPS 2025 · similarity 0.90)
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.
