Research map / 3D vision / Gaussian splatting
MAC-Splat: Multi-Attribute Consistency for High-Fidelity Sparse-View Reconstruction
Jinqian Yang, Yichen Wu, Wanhua Li, Haokun Lin, Renzhen Wang, Xiangchu Feng, Xixi Jia
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 MAC loss (ablation) | MAC-Splat · ScanNet++ Very Wide | PSNR: 17.12 -> 21.34 | Improved |
| add MAC loss without DINOv3 fusion (ablation) | MAC-Splat backbone · ScanNet++ Very Wide | PSNR: 17.12 -> 20.23 | Improved |
| add Residual Semantic Fusion with DINOv3 (ablation) | MAC-Splat · ScanNet++ Very Wide | PSNR: 20.23 -> 21.34 | Improved |
| add Residual Semantic Fusion with DINOv3 (ablation) | MAC-Splat · ScanNet++ Close | PSNR: 20.57 -> 22.76 | Improved |
| add Residual Semantic Fusion with DINOv3 (ablation) | MAC-Splat · ScanNet++ Close | LPIPS: 0.237 -> 0.146 | Improved |
| add MAC loss (ablation) | MAC-Splat · ScanNet++ Close | PSNR: 19.94 -> 22.76 | Improved |
| add MAC loss (ablation) | MAC-Splat · ScanNet++ Wide | LPIPS: 0.186 -> 0.142 | Improved |
| add DINOv3 descriptor fusion (ablation) | MAC-Splat · ScanNet++ (ablation) | LPIPS: ~22% | Improved |
| MAC-Splat vs Splatt3R (adds MAC loss and fused correspondences) | Splatt3R · ScanNet++ (average of four splits) | PSNR: more than 4.5 dB | Improved |
| MAC-Splat vs Splatt3R | Splatt3R · ScanNet++ (average of four splits) | LPIPS: about 44% | Improved |
| wider baseline (Close to Very Wide) for MAC-Splat | MAC-Splat · ScanNet++ | PSNR: 22.76 -> 21.34 | No clear effect |
| wider baseline (Close to Very Wide) for PixelSplat | PixelSplat · ScanNet++ | PSNR: 23.98 -> 15.36 | Hurt |
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.
- PixelSplat — pixelSplat: 3D Gaussian Splats from Image Pairs for Scalable Generalizable 3D Reconstruction (CVPR 2024) · better on 4 of 5 shared results
- MVSplat — MVSplat: Efficient 3D Gaussian Splatting from Sparse Multi-View Images (ECCV 2024) · better on 5 of 5 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
- MCGS: Multiview Consistency Enhancement for Sparse-View 3D Gaussian Radiance Fields (TPAMI 2025 · similarity 0.92)
- Quantifying and Alleviating Co-Adaptation in Sparse-View 3D Gaussian Splatting (NeurIPS 2025 · similarity 0.91)
- Binocular-Guided 3D Gaussian Splatting with View Consistency for Sparse View Synthesis (NeurIPS 2024 · similarity 0.91)
- Generative Sparse-View Gaussian Splatting (CVPR 2025 · similarity 0.91)
- PF3plat: Pose-Free Feed-Forward 3D Gaussian Splatting for Novel View Synthesis (ICML 2025 · similarity 0.90)
- Learning Efficient Fuse-and-Refine for Feed-Forward 3D Gaussian Splatting (NeurIPS 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.
