Research map / Image generation / Image restoration
Gain-MLP: Improving HDR Gain Map Encoding via a Lightweight MLP
Trevor D. Canham, SaiKiran Tedla, Michael J. Murdoch, Michael S. Brown
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 |
|---|---|---|---|
| replace JPEG gain map with lightweight MLP encoding | Gain map encoding · Cyriac + Chen pooled (gain map encoding) | PSNR: 38.29 -> 47.60 | Improved |
| use exponential residual (gamma map) instead of multiplicative gain map (ablation) | JPEG · Cyriac + Chen pooled (gain map encoding) | PSNR: 38.29 -> 41.45 | Improved |
| use gamma map instead of gain map (ablation) | HEIC · Cyriac + Chen pooled (gain map encoding) | PSNR: 39.20 -> 42.21 | Improved |
| use gamma map instead of gain map (ablation) | JPEG-XL · Cyriac + Chen pooled (gain map encoding) | PSNR: 37.65 -> 40.58 | Improved |
| use gamma map instead of gain map (ablation) | MLP · Cyriac + Chen pooled (gain map encoding) | PSNR: 47.60 -> 48.53 | Improved |
| use gamma map instead of gain map (ablation) | MLP · Cyriac + Chen pooled (gain map encoding) | Delta E00: 1.02 -> 0.78 | Improved |
| add SDR RGB values to position as MLP input (ablation) | Gamma-MLP · Cyriac manual tone mapping (Gamma-MLP 2x16) | PSNR: 36.6 -> 51.4 | Improved |
| add SDR RGB values to position as MLP input (ablation) | Gamma-MLP · Cyriac manual tone mapping (Gamma-MLP 2x16) | HDR-VDP3: 7.95 -> 9.74 | Hurt |
| replace Gain-MLP with Gamma-MLP versus MLP-ITM at one third the size | Gamma-MLP · Cyriac + Chen pooled (gain map encoding) | PSNR: 47.25 -> 48.53 | Improved |
| use Gamma-MLP instead of MLP-ITM | Gamma-MLP · Cyriac + Chen pooled (gain map encoding) | HDR-VDP3: 9.13 -> 9.11 | No clear effect |
| use Gain-MLP instead of Direct-MLP | Gain-MLP · Cyriac manual TM, low optimization time | PSNR | Mixed |
| use Gamma-MLP on Chen et al. YouTube TM instead of MLP-ITM | Gamma-MLP · Chen et al. UHD (YouTube TM) | PSNR: 41.6 -> 41.5 | No clear effect |
Closest papers by meaning
- Learned HDR Image Compression for Perceptually Optimal Storage and Display (ECCV 2024 · similarity 0.88)
- Dual-Output Multi-Exposure HDR Reconstruction via SDR Fusion and Gain Map Inverse Tone Mapping (ECCV 2026 · similarity 0.87)
- HDR Image Generation via Gain Map Decomposed Diffusion (ICCV 2025 · similarity 0.86)
- Learning Gain Map for Inverse Tone Mapping (ICLR 2025 · similarity 0.86)
- GMODiff: One-Step Gain Map Refinement with Diffusion Priors for Efficient HDR Reconstruction (ECCV 2026 · similarity 0.86)
- Zero-Shot Structure-Preserving Diffusion Model for High Dynamic Range Tone Mapping (CVPR 2024 · similarity 0.84)
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.
