Research map / Video / Video generation
Tempo-SAM3D: Monocular Video to 4D via Temporal Memory-Guided Generation
Baicheng Li, Dong Wu, Yingdian Cao, Haoxiang Yang, Yiwen Lu, Zike Yan, Hongbin Zha
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 self-attention memory (ablation) | SAM3D · Tempo-SAM3D dataset | FVD: 1218.63 -> 1076.52 | Improved |
| add self-attention memory (ablation) | SAM3D · Tempo-SAM3D dataset | LPIPS: 0.128 -> 0.131 | No clear effect |
| add self-attention memory (ablation) | SAM3D · Tempo-SAM3D dataset | ACC5cm: 90.3 -> 89.8 | No clear effect |
| add temporal guidance (ablation) | SAM3D + self-attn memory · Tempo-SAM3D dataset | FVD: 1076.52 -> 935.18 | Improved |
| add alignment guidance (ablation) | SAM3D + memory + temporal guidance · Tempo-SAM3D dataset | FVD: 935.18 -> 871.25 | Improved |
| add alignment guidance (ablation) | SAM3D + memory + temporal guidance · Tempo-SAM3D dataset | IoU: 73.3 -> 82.6 | Improved |
| add alignment guidance (ablation) | SAM3D + memory + temporal guidance · Tempo-SAM3D dataset | ACC5cm: 90.1 -> 95.2 | Improved |
| add HexPlane temporal smoothing (ablation) | Tempo-SAM3D w/o HexPlane · Tempo-SAM3D dataset | FVD: 871.25 -> 732.45 | Improved |
| add HexPlane temporal smoothing (ablation) | Tempo-SAM3D w/o HexPlane · Tempo-SAM3D dataset | PSNR: 20.18 -> 20.12 | No clear effect |
| add cross-attention visual memory (ablation) | Tempo-SAM3D · construction worker sequence | appearance consistency | Improved |
| use Tempo-SAM3D instead of V2M4 | V2M4 benchmark · Simple (Consistent4D) | FVD: 825.59 -> 841.27 | Hurt |
| use Tempo-SAM3D instead of V2M4 | V2M4 benchmark · Complex (Mixamo & Sketchfab) | FVD: 666.04 -> 642.38 | Improved |
Closest papers by meaning
- DreamScene4D: Dynamic Multi-Object Scene Generation from Monocular Videos (NeurIPS 2024 · similarity 0.89)
- Diffusion4D: Fast Spatial-temporal Consistent 4D generation via Video Diffusion Models (NeurIPS 2024 · similarity 0.89)
- 4Diffusion: Multi-view Video Diffusion Model for 4D Generation (NeurIPS 2024 · similarity 0.89)
- SV4D: Dynamic 3D Content Generation with Multi-Frame and Multi-View Consistency (ICLR 2025 · similarity 0.89)
- 4Real: Towards Photorealistic 4D Scene Generation via Video Diffusion Models (NeurIPS 2024 · 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.
