Research map / 3D vision / Gaussian splatting
SunFaded: Illumination-Aware Gaussian Splatting for Dark Scenes with Camera-Mounted Active Lighting
Wenjie Chang, Tianle Ding, Wenfei Yang, Tianzhu Zhang
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 geometry prior supervision (ablation) | SunFaded · DarkRobotic average | PSNR: 40.92 -> 41.02 | Improved |
| add geometry prior supervision (ablation) | SunFaded · DarkRobotic average | LPIPS: 0.0339 -> 0.0286 | Improved |
| increase beta from 0.1 to 10 in illumination-weighted loss (ablation) | SunFaded · DarkRobotic average | PSNR: 33.16 -> 41.02 | Improved |
| increase beta from 10 to 50 (ablation) | SunFaded · DarkRobotic average | PSNR: 41.02 -> 37.03 | Hurt |
| shrink lighting tile from 16x16 to 4x4 (ablation) | SunFaded · DarkRobotic average | PSNR: 41.02 -> 41.25 | No clear effect |
| shrink lighting tile from 16x16 to 4x4 (ablation) | SunFaded · DarkRobotic average | FPS: 193 -> 95 | Hurt |
| enlarge lighting tile from 16x16 to 64x64 (ablation) | SunFaded · DarkRobotic average | PSNR: 41.02 -> 33.55 | Hurt |
| use image-space tile lighting module instead of per-Gaussian appearance | SunFaded vs GS-W · DarkRobotic average | PSNR: 35.62 -> 41.02 | Improved |
| use SunFaded without light calibration vs DarkGS | SunFaded · DarkRobotic average | PSNR: 35.50 -> 41.02 | Improved |
| use SunFaded vs DarkGS on new phone-flash capture | SunFaded · DarkPhone | reconstruction | Improved |
| train with SunFaded instead of GS-W | SunFaded · DarkRobotic | training time: ~287mins -> ~22mins | Improved |
| apply SunFaded vs GS-W on DarkPhone Bucket LPIPS | SunFaded · DarkPhone Bucket | LPIPS: 0.3132 -> 0.1075 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- WildGaussians — WildGaussians: 3D Gaussian Splatting In the Wild (NeurIPS 2024) · better on 3 of 3 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
- Ref-GS: Directional Factorization for 2D Gaussian Splatting (CVPR 2025 · similarity 0.90)
- GI-GS: Global Illumination Decomposition on Gaussian Splatting for Inverse Rendering (ICLR 2025 · similarity 0.90)
- Decoupling Appearance Variations with 3D Consistent Features in Gaussian Splatting (AAAI 2025 · similarity 0.89)
- FewViewGS: Gaussian Splatting with Few View Matching and Multi-stage Training (NeurIPS 2024 · similarity 0.89)
- Structure Consistent Gaussian Splatting with Matching Prior for Few-shot Novel View Synthesis (NeurIPS 2024 · similarity 0.89)
- IBGS: Image-Based Gaussian Splatting (NeurIPS 2025 · similarity 0.89)
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.
