Research map / 3D vision / Gaussian splatting
MVFusion-GS: Motion-Variance Guided Temporal Attention for High-Quality Dynamic Gaussian Splatting
Jianwei Hu, Tingxuan Huang, Hengyu Zhou, Ningna Wang, Xiaohu Guo, Jinshan Lai, Bin Wang
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 motion-variance guided refinement and MotionFormer temporal attention | DeGauss · NeRF On-the-go background | LPIPS: 0.113 -> 0.085 | Improved |
| add MVG+MFTA | DeGauss · NeRF On-the-go foreground | PSNR: 22.93 -> 27.60 | Improved |
| add MVG+MFTA | DeGauss · RobustNeRF | PSNR: 28.89 -> 29.39 | Improved |
| add MVG+MFTA | DeGauss · Neu3D | PSNR: 31.52 -> 32.07 | Improved |
| add MVG+MFTA plug-in to 4DGS without decomposition | 4DGS · Neu3D | PSNR: 31.12 -> 31.66 | Improved |
| add motion statistics (global trajectory signature) with position variance only (ablation) | MVFusion-GS · Neu3D | PSNR: 31.52 -> 31.93 | Improved |
| use full 13D signature instead of position-only variance (ablation) | MVFusion-GS · Neu3D | PSNR: 31.93 -> 32.07 | Improved |
| add MFTA temporal attention (ablation) | MVFusion-GS · Neu3D | PSNR: 31.78 -> 32.07 | Improved |
| use query-centered cross-attention instead of self-attention (ablation) | MFTA · Neu3D | PSNR: 32.01 -> 32.07 | Improved |
| add MVG+MFTA refinement (ablation) | DeGauss · NeRF On-the-go background | PSNR: 23.86 -> 23.94 | No clear effect |
| add MVG (ablation) | MVFusion-GS · Neu3D | SSIM: w/o MVG SSIM 0.937 vs base 0.942 vs full 0.943 | Mixed |
| use 0.7/0.2/0.1 weighted motion score instead of uniform weights (ablation) | MVG · Neu3D | PSNR: 31.88 -> 32.07 | 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.
- DeSplat — DeSplat: Decomposed Gaussian Splatting for Distractor-Free Rendering (CVPR 2025) · better on 2 of 2 shared results
- MangoGS — Mango-GS: Enhancing Spatio-Temporal Consistency in Dynamic Scenes Reconstruction using Multi-Frame Node-Guided 4D Gaussian Splatting (ICLR 2026) · 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
- HAIF-GS: Hierarchical and Induced Flow-Guided Gaussian Splatting for Dynamic Scene (NeurIPS 2025 · similarity 0.91)
- Motion Decoupled 3D Gaussian Splatting for Dynamic Object Representation (AAAI 2025 · similarity 0.91)
- SC-GS: Sparse-Controlled Gaussian Splatting for Editable Dynamic Scenes (CVPR 2024 · similarity 0.91)
- MoDec-GS: Global-to-Local Motion Decomposition and Temporal Interval Adjustment for Compact Dynamic 3D Gaussian Splatting (CVPR 2025 · similarity 0.90)
- MotionGS: Exploring Explicit Motion Guidance for Deformable 3D Gaussian Splatting (NeurIPS 2024 · similarity 0.90)
- Gaussian Splatting Lucas-Kanade (ICLR 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.
