Research map / 3D vision / Gaussian splatting
TimeFormer: Capturing Temporal Relationships of Deformable 3D Gaussians for Robust Reconstruction
Dadong Jiang, Zhi Hou, Zhihui Ke, Xianghui Yang, Xiaobo Zhou, Tie Qiu
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 TimeFormer to 4DGS | 4DGS · N3DV (multi-view, mean of 6 scenes) | PSNR: 31.10 -> 31.84 | Improved |
| add TimeFormer to STGS | STGS · N3DV (multi-view, mean of 6 scenes) | PSNR: 31.62 -> 32.23 | Improved |
| add TimeFormer to 4DGS | 4DGS · HyperNeRF (monocular, mean of 6 scenes) | PSNR: 27.14 -> 28.03 | Improved |
| add TimeFormer to DeformGS | DeformGS · HyperNeRF (monocular, mean of 6 scenes) | PSNR: 26.09 -> 27.03 | Improved |
| add TimeFormer to DeformGS | DeformGS · NeRF-DS (monocular, mean of 6 scenes) | PSNR: 23.59 -> 24.10 | Improved |
| add TimeFormer to DeformGS (inference without TimeFormer) | DeformGS · HyperNeRF (monocular) | FPS: 30.1 -> 58.9 | Improved |
| add TimeFormer to 4DGS | 4DGS · N3DV (multi-view) | FPS: 31.7 -> 37.9 | Improved |
| add TimeFormer to STGS | STGS · N3DV (multi-view) | training time: 41 min -> 78 min | Hurt |
| share weights of the two deformation fields (add sharing) (ablation) | TimeFormer + DeformGS · NeRF-DS (3 scenes: press, sieve, bell) ablation | PSNR: 24.81 -> 26.04 | Improved |
| use random instead of continuous timestamp sampling (ablation) | TimeFormer + DeformGS · NeRF-DS (3 scenes: press, sieve, bell) ablation | PSNR: 25.78 -> 26.06 | Improved |
| increase encoder layers M from 1 to 4 (ablation) | TimeFormer + DeformGS · NeRF-DS (3 scenes: press, sieve, bell) ablation | PSNR: 25.96 -> 26.04 | No clear effect |
| increase time batch B from 2 to 4 (random sampling) (ablation) | TimeFormer + DeformGS · NeRF-DS (3 scenes: press, sieve, bell) ablation | PSNR: 25.89 -> 26.06 | No clear effect |
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.
- NeRF-DS — NeRF-DS: Neural Radiance Fields for Dynamic Specular Objects (CVPR 2023) · 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
- Deformable 3D Gaussians for High-Fidelity Monocular Dynamic Scene Reconstruction (CVPR 2024 · similarity 0.92)
- Motion Decoupled 3D Gaussian Splatting for Dynamic Object Representation (AAAI 2025 · similarity 0.90)
- FreeTimeGS: Free Gaussian Primitives at Anytime Anywhere for Dynamic Scene Reconstruction (CVPR 2025 · similarity 0.89)
- Gaussian-Flow: 4D Reconstruction with Dynamic 3D Gaussian Particle (CVPR 2024 · similarity 0.89)
- DGS-LRM: Real-Time Deformable 3D Gaussian Reconstruction From Monocular Videos (NeurIPS 2025 · similarity 0.89)
- MotionGS: Exploring Explicit Motion Guidance for Deformable 3D Gaussian Splatting (NeurIPS 2024 · 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.
