Research map / Image generation / Image restoration
FastGaMer: Efficient GainMap Learning for Practical Inverse Tone Mapping
Yuanshen Guan, Ruikang Xu, Chang Chen, Yinuo Liao, Dehua Song, Fenglong Song, Zhiwei Xiong
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 Gain Map learning (color GM supervision and scale estimation) instead of direct HDR regression (ablation) | FastGaMer · Real-captured test set | PQ-PSNR: 26.59 -> 30.01 | Improved |
| add dynamic bilateral grid slicing (ablation) | FastGaMer · Real-captured test set | PQ-PSNR: 29.68 -> 30.01 | Improved |
| add learnable 3D LUT transform (ablation) | FastGaMer · Real-captured test set | PQ-PSNR: 29.75 -> 30.01 | Improved |
| add lightweight neural modulation (ablation) | FastGaMer · Real-captured test set | PQ-PSNR: 29.79 -> 30.01 | Improved |
| downsample 4K input SDR to 2K (ablation) | FastGaMer · Real-captured test set | PQ-PSNR: 30.01 -> 29.98 | No clear effect |
| downsample 4K input SDR to 1K (ablation) | FastGaMer · Real-captured test set | PQ-PSNR: 30.01 -> 29.57 | Hurt |
| downsample 4K input SDR to 2K (ablation) | FastGaMer · Synthetic test set | PQ-PSNR: 32.06 -> 32.21 | Improved |
| predict color gain map with grids and LUTs instead of ITMLUT | ITMLUT · Synthetic test set | PQ-PSNR: 30.66 -> 32.06 | Improved |
| FastGaMer vs ITMLUT runtime at 4K | ITMLUT · 4K runtime | runtime (ms): 18.5 -> 6.20 | Improved |
| FastGaMer vs ITMLUT on linear-domain PSNR on real set | ITMLUT · Real-captured test set | Linear PSNR: 32.61 -> 31.28 | Hurt |
| FastGaMer vs GMNet on HDRVDP3 | GMNet · Synthetic test set | HDRVDP3: 9.385 -> 9.263 | Hurt |
Closest papers by meaning
- Learning Differential Pyramid Representation for Tone Mapping (NeurIPS 2025 · similarity 0.87)
- Zero-Shot Structure-Preserving Diffusion Model for High Dynamic Range Tone Mapping (CVPR 2024 · similarity 0.86)
- Plug-and-Play Tri-Branch Invertible Block for Image Rescaling (AAAI 2025 · similarity 0.86)
- GaussHDR: High Dynamic Range Gaussian Splatting via Learning Unified 3D and 2D Local Tone Mapping (CVPR 2025 · similarity 0.86)
- PixelART: Image-to-Layer Decomposition without Latents or Text-to-Image Pretraining (NeurIPS 2026 · similarity 0.84)
- Diffusion Reflectance Map: Single-Image Stochastic Inverse Rendering of Illumination and Reflectance (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.
