Research map / 3D vision / Gaussian splatting
GaSLight: Gaussian Splats for Spatially-Varying Lighting in HDR
Christophe Bolduc, Yannick Hold-Geoffroy, Jean-François Lalonde
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 |
|---|---|---|---|
| use diffusion-based multi-exposure HDR estimator instead of LaNet within the Gaussian splat pipeline (ablation) | GaSLight · PanoHDR-NeRF Chess room | Relight PSNR: 23.00 -> 30.50 | Improved |
| replace NeRF/LANet lighting with HDR Gaussian splats plus diffusion HDR estimation | GaSLight vs PanoHDR-NeRF · PanoHDR-NeRF Chess room | Relight MSE: 0.0134 -> 0.0026 | Improved |
| use GaSLight HDR estimator versus prior LDR-to-HDR methods | GaSLight · Si-HDR (113 unsaturated) | Relight MSE: 0.0178 -> 0.0131 (vs SingleHDR) | Improved |
| GaSLight versus prior LDR-to-HDR methods on BtP-HDR | GaSLight · BtP-HDR | Relight PSNR: best MSE 0.0041, but PSNR 28.7022 vs MaskHDR 30.9510 | Mixed |
| general (non-category-specific) network versus GlowGAN trained on landscapes | GaSLight · Si-HDR landscape subset (23 images) | Relight PSNR: 23.82 -> 25.05 | Improved |
| reconstruct HDR 3DGS with spherical harmonics colors on inconsistent per-image HDR predictions (ablation) | GaSLight · multi-view consistency (qualitative) | multi-view consistency | Improved |
Closest papers by meaning
- Lighting Every Darkness with 3DGS: Fast Training and Real-Time Rendering for HDR View Synthesis (NeurIPS 2024 · similarity 0.89)
- GI-GS: Global Illumination Decomposition on Gaussian Splatting for Inverse Rendering (ICLR 2025 · similarity 0.88)
- IBGS: Image-Based Gaussian Splatting (NeurIPS 2025 · similarity 0.88)
- DiffusionLight: Light Probes for Free by Painting a Chrome Ball (CVPR 2024 · similarity 0.87)
- pixelSplat: 3D Gaussian Splats from Image Pairs for Scalable Generalizable 3D Reconstruction (CVPR 2024 · similarity 0.87)
- RNG: Relightable Neural Gaussians (CVPR 2025 · similarity 0.87)
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.
