Research map / 3D vision / 3D generation
StyleLight: HDR Panorama Generation for Lighting Estimation and Editing
Guangcong Wang, Yinuo Yang, Chen Change Loy, Ziwei Liu
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 focal-masked loss to GAN inversion (masked to focal-masked) (ablation) | StyleLight · Laval Indoor HDR (mirror sphere) | RMSE: 0.57 -> 0.56 | Improved |
| add focal-masked loss to GAN inversion (ablation) | StyleLight · Laval Indoor HDR (matte silver sphere) | RMSE: 0.31 -> 0.30 | Improved |
| add focal-masked loss to GAN inversion (ablation) | StyleLight · Laval Indoor HDR (diffuse sphere) | RMSE: 0.16 -> 0.15 | Improved |
| add focal-masked loss to GAN inversion (ablation) | StyleLight · Laval Indoor HDR (mirror sphere) | si-RMSE: 0.56 -> 0.55 | Improved |
| add focal-masked loss to GAN inversion (ablation) | StyleLight · Laval Indoor HDR (matte silver sphere) | si-RMSE: 0.30 -> 0.29 | Improved |
| add focal-masked loss to GAN inversion (ablation) | StyleLight · Laval Indoor HDR (diffuse sphere) | si-RMSE: 0.12 -> 0.11 | Improved |
| add a separate LDR discriminator and LDR branch (coupled dual-StyleGAN) versus HDR-only with one discriminator (ablation) | StyleGAN2 panorama synthesis · LDR panorama quality (qualitative) | visual quality of LDR panoramas | Improved |
| replace regression/multi-stage lighting estimation (EMlight) with StyleLight | lighting estimation · Laval Indoor HDR test | FID: 263.8 -> 137.7 | Improved |
| replace Gardner2019 (depth annotations) with StyleLight, no annotations | lighting estimation · Laval Indoor HDR (mirror sphere) | Angular Error: 4.63 -> 4.30 | Improved |
| replace EMlight with StyleLight | lighting estimation · Laval Indoor HDR (mirror sphere) | RMSE: 0.59 -> 0.56 | Improved |
Later papers that compared with this method
1 later accepted paper reports StyleLight as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Photorealistic Object Insertion with Diffusion-Guided Inverse Rendering (ECCV 2024 · 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 1 of 1 on most shared results.
- EMlight — EMLight: Lighting Estimation via Spherical Distribution Approximation (AAAI 2021) · better on 10 of 10 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
- LuxDiT: Lighting Estimation with Video Diffusion Transformer (NeurIPS 2025 · similarity 0.87)
- Pano-NeRF: Synthesizing High Dynamic Range Novel Views with Geometry from Sparse Low Dynamic Range Panoramic Images (AAAI 2024 · similarity 0.86)
- HORIZON: High-Resolution Semantically Controlled Panorama Synthesis (AAAI 2024 · similarity 0.85)
- Do Image Editing Models Understand Lighting? (NeurIPS 2026 · similarity 0.85)
- High Dynamic Range Novel View Synthesis with Single Exposure (ICML 2025 · similarity 0.85)
- IRIS: Inverse Rendering of Indoor Scenes from Low Dynamic Range Images (CVPR 2025 · similarity 0.85)
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.
