Research map / Image generation / GANs
Unpaired Image Enhancement Featuring Reinforcement-Learning-Controlled Image Editing Software
Satoshi Kosugi, Toshihiko Yamasaki
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 RL-controlled Lightroom filters instead of differentiable filters (ablation) | Proposed method · MIT-Adobe 5K | PSNR | Improved |
| Use RL-controlled Lightroom filters instead of differentiable filters (ablation) | Proposed method · MIT-Adobe 5K | SSIM | Improved |
| Add an MSE penalty to the reward (ablation) | Proposed method · MIT-Adobe 5K | PSNR | Improved |
| Add an MSE penalty to the reward (ablation) | Proposed method · MIT-Adobe 5K | SSIM | Improved |
| Use 1D convolutional layers instead of a fully connected layer (ablation) | Proposed method · MIT-Adobe 5K | PSNR | Improved |
| Use 1D convolutional layers instead of a fully connected layer (ablation) | Proposed method · MIT-Adobe 5K | SSIM | Improved |
| Use the proposed RL-controlled editing method rather than the compared unpaired enhancement methods | Photo enhancement · MIT-Adobe 5K, 512px and 2048px | PSNR | Improved |
| Use the proposed RL-controlled editing method rather than the compared unpaired enhancement methods | Photo enhancement · MIT-Adobe 5K, 512px and 2048px | SSIM | Improved |
| Use the proposed method rather than existing methods | Photo enhancement · MIT-Adobe 5K user study | five-grade rating | Improved |
| Use the proposed method rather than existing methods | Face beautification · SCUT-FBP5500 user study | naturality and preference ratings | Improved |
Closest papers by meaning
- Deep Photo Enhancer: Unpaired Learning for Image Enhancement From Photographs With GANs (CVPR 2018 · similarity 0.90)
- Global and Local Enhancement Networks for Paired and Unpaired Image Enhancement (ECCV 2020 · similarity 0.89)
- Distort-and-Recover: Color Enhancement Using Deep Reinforcement Learning (CVPR 2018 · similarity 0.86)
- Aesthetics-Guided Low-Light Enhancement (TPAMI 2025 · similarity 0.85)
- SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equations (ICLR 2022 · similarity 0.85)
- E$^2$GAN: Efficient Training of Efficient GANs for Image-to-Image Translation (ICML 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.
