Research map / 3D vision / NeRF and radiance fields
LuxRemix: Lighting Decomposition and Remixing for Indoor Scenes
Ruofan Liang, Norman Müller, Ethan Weber, Duncan Zauss, Nandita Vijaykumar, Peter Kontschieder, Christian Richardt
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 |
|---|---|---|---|
| channel-wise mask token addition instead of side-by-side token concatenation (ablation) | LuxRemix-SV · Single-image decomposition | PSNR: 25.20 -> 27.68 | Improved |
| FLUX DiT with LoRA instead of SD U-Net (ablation) | LuxRemix-SV · Single-image decomposition | PSNR: 27.13 -> 27.68 | Improved |
| FLUX DiT with LoRA instead of SD U-Net (ablation) | LuxRemix-SV · Single-image decomposition | SSIM: 0.857 -> 0.898 | Improved |
| channel-wise mask addition instead of FLUX token concatenation (ablation) | LuxRemix-SV · Single-image decomposition | SSIM: 0.865 -> 0.898 | Improved |
| fine-tune for per-light decomposition vs general editor Qwen-Image | LuxRemix-SV · Single-image decomposition | PSNR: 18.23 -> 27.68 | Improved |
| add multi-view harmonization instead of per-view processing (ablation) | LuxRemix-MV · Multi-view harmonization | PSNR: 25.14 -> 30.76 | Improved |
| harmonize from sparse reference views instead of mask-guided multi-view editing (ablation) | LuxRemix-MV · Multi-view harmonization | PSNR: 26.37 -> 30.76 | Improved |
| mask-guided multi-view editing vs per-view (ablation) | LuxRemix-MV-Edit · Multi-view harmonization | SSIM: 0.807 -> 0.794 | Hurt |
Closest papers by meaning
- Decoupled Illumination Priors for Spatially Controllable Multi-View Indoor Scene Relighting (ECCV 2026 · similarity 0.91)
- Physically-Based Editing of Indoor Scene Lighting from a Single Image (ECCV 2022 · similarity 0.90)
- ScribbleLight: Single Image Indoor Relighting with Scribbles (CVPR 2025 · similarity 0.89)
- LumiNet: Latent Intrinsics Meets Diffusion Models for Indoor Scene Relighting (CVPR 2025 · similarity 0.89)
- Do Image Editing Models Understand Lighting? (NeurIPS 2026 · similarity 0.88)
- SyncLight: Single-Edit Multi-View Relighting (NeurIPS 2026 · similarity 0.86)
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.
