Research map / 3D vision / 3D generation
4DGS360: 360° Gaussian Reconstruction of Dynamic Objects from a Single Video
Jae Won Jang, Yeonjin Chang, Wonsik Shin, Juhwan Cho, Nojun Kwak
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 |
|---|---|---|---|
| replace HiMoR 2D-track initialization with AnchorTAP3D initialization (4DGS360) | 4DGS360 · iPhone360 (6 scenes, extreme views) | CLIP-I / LPIPS: Block2 CLIP-I 0.8422 -> 0.9021; Goat LPIPS 0.3260 -> 0.2244 | Improved |
| 4DGS360 vs HiMoR on iPhone | 4DGS360 · iPhone dataset (mean) | LPIPS: 0.4639 -> 0.3877 | Improved |
| 4DGS360 vs HiMoR on CLIP-T | 4DGS360 · iPhone360 Goat/Jacket/Jelly | CLIP-T: Goat 0.9426 -> 0.9345; Jacket 0.9526 -> 0.9445 | Mixed |
| initialize with naive 3D tracker TAPIP3D instead of anchor-guided AnchorTAP3D (w/o Anchor) (ablation) | 4DGS360 · iPhone360 walk-around, jelly | qualitative 360 reconstruction | Hurt |
| initialize from unprojected 2D tracks instead of 3D tracking (w/o 3D init) (ablation) | 4DGS360 · iPhone360 walk-around, jelly | qualitative 360 reconstruction | Hurt |
| anchor confidence threshold tau 0.5 vs 0.7 vs 0.3 (ablation) | AnchorTAP3D · iPhone360 goat, jelly | qualitative rendering | Mixed |
| use MoGe-only depth without LiDAR correction (ablation) | 4DGS360 · iPhone360 jelly | qualitative alignment | Mixed |
| combine with DIFIX3D+ diffusion prior | 4DGS360 · iPhone360 | qualitative/quantitative | 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.
- HiMoR — HiMoR: Monocular Deformable Gaussian Reconstruction with Hierarchical Motion Representation (CVPR 2025) · better on 12 of 12 shared results
- MoSca — MoSca: Dynamic Gaussian Fusion from Casual Videos via 4D Motion Scaffolds (CVPR 2025) · better on 10 of 10 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
- DGS-LRM: Real-Time Deformable 3D Gaussian Reconstruction From Monocular Videos (NeurIPS 2025 · similarity 0.90)
- World from Motion: Generative Dynamic Gaussian Reconstruction from Monocular Video (NeurIPS 2026 · similarity 0.89)
- 4D Gaussian Splatting in the Wild with Uncertainty-Aware Regularization (NeurIPS 2024 · similarity 0.89)
- 4D3R: Motion-Aware Neural Reconstruction and Rendering of Dynamic Scenes from Monocular Videos (NeurIPS 2025 · similarity 0.88)
- Instant4D: 4D Gaussian Splatting in Minutes (NeurIPS 2025 · similarity 0.88)
- Restage4D: Reanimating Deformable 3D Reconstruction from a Single Video (NeurIPS 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.
