Research map / 3D vision / Gaussian splatting
BAGS: Blur Agnostic Gaussian Splatting through Multi-Scale Kernel Modeling
Cheng Peng, Yutao Tang, Yifan Zhou, Nengyu Wang, Xijun Liu, Deming Li, Rama Chellappa
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 Blur Proposal Network with coarse-to-fine kernel optimization to Mip-Splatting | Mip-Splatting · Deblur-NeRF camera motion blur | PSNR: 21.87 -> 26.70 | Improved |
| add BAGS to Mip-Splatting on defocus blur | Mip-Splatting · Deblur-NeRF defocus blur | PSNR: 21.08 -> 23.95 | Improved |
| add BAGS on mixed-resolution unbounded scenes | Mip-Splatting · Mip-NeRF 360 mixed resolution | PSNR: 26.11 -> 27.11 | Improved |
| add BAGS on low-light drone motion blur | Mip-Splatting · Low-light drone | PSNR: 26.50 -> 29.86 | Improved |
| pre-restore images with HAT super-resolution before Mip-Splatting | Mip-Splatting · Mip-NeRF 360 mixed resolution | PSNR: 26.11 -> 24.99 | Hurt |
| pre-restore images with MPR deblurring before Mip-Splatting | Mip-Splatting · Low-light drone | PSNR: 26.50 -> 26.36 | No clear effect |
| add RGBD features to the Blur Proposal Network (ablation) | BAGS · camera motion and defocus blur | PSNR: +1.57/0.92 dB PSNR, +0.03/0.03 SSIM | Improved |
| add coarse-to-fine kernel optimization (ablation) | BAGS · defocus blur | PSNR: +1.64 dB PSNR, +0.08 SSIM | Improved |
| use dense full kernel instead of sparse deformable kernel (ablation) | BAGS · camera motion blur | PSNR/SSIM | Improved |
| increase final kernel size K1 (ablation) | BAGS · motion/defocus and low-light scenarios | reconstruction quality | Improved |
| increase final kernel size K1 (ablation) | BAGS · mixed resolution | reconstruction quality | Hurt |
| use coarse-to-fine training instead of single-scale training (ablation) | BAGS · training time | training time: under fifteen minutes vs two hours for PDRF | Improved |
Later papers that compared with this method
6 later accepted papers report BAGS as a baseline in their results tables; 5 of them report a better number on most of the results they share with it.
- Unblur-SLAM: Dense Neural SLAM for Blurry Inputs (CVPR 2026 · better on 1 of 1 shared result)
- CoCoGaussian: Leveraging Circle of Confusion for Gaussian Splatting from Defocused Images (CVPR 2025 · better on 8 of 8 shared results)
- DoF-Gaussian: Controllable Depth-of-Field for 3D Gaussian Splatting (CVPR 2025 · better on 3 of 6 shared results)
- DOF-GS: Adjustable Depth-of-Field 3D Gaussian Splatting for Post-Capture Refocusing, Defocus Rendering and Blur Removal (CVPR 2025 · better on 4 of 4 shared results)
- Exploiting Deblurring Networks for Radiance Fields (CVPR 2025 · better on 2 of 2 shared results)
- CoMoGaussian: Continuous Motion-Aware Gaussian Splatting from Motion-Blurred Images (ICCV 2025 · better on 2 of 2 shared results)
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.
- Mip-Splatting — Mip-Splatting: Alias-free 3D Gaussian Splatting (CVPR 2024) · better on 6 of 6 shared results
- DP-NeRF — DP-NeRF: Deblurred Neural Radiance Field With Physical Scene Priors (CVPR 2023) · better on 3 of 3 shared results
- Mip-NeRF 360 — Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields (CVPR 2022) · better on 2 of 2 shared results
- Deblur-NeRF — Deblur-NeRF: Neural Radiance Fields From Blurry Images (CVPR 2022) · 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
- Dynamic Gaussian Splatting from Defocused and Motion-blurred Monocular Videos (NeurIPS 2025 · similarity 0.90)
- IBGS: Image-Based Gaussian Splatting (NeurIPS 2025 · similarity 0.89)
- Generative Sparse-View Gaussian Splatting (CVPR 2025 · similarity 0.88)
- Textured Gaussians for Enhanced 3D Scene Appearance Modeling (CVPR 2025 · similarity 0.88)
- GS-Blur: A 3D Scene-Based Dataset for Realistic Image Deblurring (NeurIPS 2024 · similarity 0.88)
- Compressed 3D Gaussian Splatting for Accelerated Novel View Synthesis (CVPR 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.
