Research map / 3D vision / Gaussian splatting
ReSplat: Learning Recurrent Gaussian Splatting
Haofei Xu, Daniel Barath, Andreas Geiger, Marc Pollefeys
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 rendering error feedback to the recurrent model (ablation) | ReSplat-Small · DL3DV (8 views, 256x448) | PSNR: 27.19 -> 29.07 | Improved |
| use feature error rather than RGB error only (ablation) | ReSplat-Small · DL3DV (8 views, 256x448) | PSNR: 27.90 -> 28.77 | Improved |
| combine RGB and feature errors by addition rather than concatenation (ablation) | ReSplat-Small · DL3DV (8 views, 256x448) | PSNR: 28.93 -> 29.07 | Improved |
| align coordinates to the middle input view instead of COLMAP frame (ablation) | ReSplat-Small · DL3DV (8 views, 256x448) | PSNR: 28.14 -> 29.07 | Improved |
| add state input to the recurrent network (ablation) | ReSplat-Small · DL3DV (8 views, 256x448) | PSNR: 27.79 -> 29.07 | Improved |
| add kNN attention to the recurrent model (ablation) | ReSplat-Small · DL3DV (8 views, 256x448) | PSNR: 28.58 -> 29.07 | Improved |
| add global attention to the recurrent model (ablation) | ReSplat-Small · DL3DV (8 views, 256x448) | PSNR: 28.96 -> 29.07 | Improved |
| add kNN attention to the initial model (ablation) | ReSplat-Small init · DL3DV (8 views, 256x448) | PSNR: 25.30 -> 26.77 | Improved |
| add global attention to the initial model (ablation) | ReSplat-Small init · DL3DV (8 views, 256x448) | PSNR: 26.33 -> 26.77 | Improved |
| use DINOv2 features instead of ResNet for rendering error (ablation) | ReSplat-Small · DL3DV (8 views, 256x448) | PSNR: 29.07 -> 29.00 | No clear effect |
| use 4x compression instead of 16x in the initial model (ablation) | ReSplat-Small init · DL3DV (8 views, 256x448) | PSNR: 26.77 -> 28.36 | Improved |
| use 64x compression instead of 16x in the initial model (ablation) | ReSplat-Small init · DL3DV (8 views, 256x448) | PSNR: 26.77 -> 24.77 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 4 of 4 on most shared results.
- DepthSplat — DepthSplat: Connecting Gaussian Splatting and Depth (CVPR 2025) · better on 3 of 3 shared results
- MVSplat — MVSplat: Efficient 3D Gaussian Splatting from Sparse Multi-View Images (ECCV 2024) · better on 2 of 2 shared results
- Long-LRM — Long-LRM: Long-sequence Large Reconstruction Model for Wide-coverage Gaussian Splats (ICCV 2025) · better on 1 of 1 shared result
- Scaffold-GS — Scaffold-GS: Structured 3D Gaussians for View-Adaptive Rendering (CVPR 2024) · 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
- pixelSplat: 3D Gaussian Splats from Image Pairs for Scalable Generalizable 3D Reconstruction (CVPR 2024 · similarity 0.93)
- Learning Efficient Fuse-and-Refine for Feed-Forward 3D Gaussian Splatting (NeurIPS 2025 · similarity 0.91)
- ReorgGS: Equivalent Distribution Reorganization for 3D Gaussian Splatting (NeurIPS 2026 · similarity 0.91)
- DropGaussian: Structural Regularization for Sparse-view Gaussian Splatting (CVPR 2025 · similarity 0.91)
- GaussianSpa: An "Optimizing-Sparsifying" Simplification Framework for Compact and High-Quality 3D Gaussian Splatting (CVPR 2025 · similarity 0.91)
- Generative Sparse-View Gaussian Splatting (CVPR 2025 · similarity 0.91)
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.
