# IRPO: Boosting Image Restoration via Post-training GRPO

arXiv: https://arxiv.org/abs/2512.00814 · HTML: https://mlsanity.com/paper/2512.00814/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 70 of its 94 references that are in the map.
- First posted on arXiv 30 Nov 2025.

## Builds on (most cited first, 30 of 70)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [LLaMA: Open and Efficient Foundation Language Models](https://mlsanity.com/paper/2302.13971/) (2023)
- [The Unreasonable Effectiveness of Deep Features as a Perceptual Metric](https://mlsanity.com/paper/1801.03924/) (2018)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://mlsanity.com/paper/2501.12948/) (2025)
- [LAION-5B: An open large-scale dataset for training next generation image-text models](https://mlsanity.com/paper/2210.08402/) (2022)
- [Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2204.05862/) (2022)
- [Perceptual Losses for Real-Time Style Transfer and Super-Resolution](https://mlsanity.com/paper/1603.08155/) (2016)
- [SwinIR: Image Restoration Using Swin Transformer](https://mlsanity.com/paper/2108.10257/) (2021)
- [Restormer: Efficient Transformer for High-Resolution Image Restoration](https://mlsanity.com/paper/2111.09881/) (2021)
- [Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising](https://mlsanity.com/paper/1608.03981/) (2016)
- [ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks](https://mlsanity.com/paper/1809.00219/) (2018)
- [Image Super-Resolution Using Deep Convolutional Networks](https://mlsanity.com/paper/1501.00092/) (2014)
- [MUSIQ: Multi-scale Image Quality Transformer](https://mlsanity.com/paper/2108.05997/) (2021)
- [Multi-Stage Progressive Image Restoration](https://mlsanity.com/paper/2102.02808/) (2021)
- [Image Quality Assessment: Unifying Structure and Texture Similarity](https://mlsanity.com/paper/2004.07728/) (2020)
- [Deep Multi-scale Convolutional Neural Network for Dynamic Scene Deblurring](https://mlsanity.com/paper/1612.02177/) (2016)
- [Deep Retinex Decomposition for Low-Light Enhancement](https://mlsanity.com/paper/1808.04560/) (2018)
- [Simple Baselines for Image Restoration](https://mlsanity.com/paper/2204.04676/) (2022)
- [The Perception-Distortion Tradeoff](https://mlsanity.com/paper/1711.06077/) (2017)
- [Noise2Noise: Learning Image Restoration without Clean Data](https://mlsanity.com/paper/1803.04189/) (2018)
- [FFDNet: Toward a Fast and Flexible Solution for CNN based Image Denoising](https://mlsanity.com/paper/1710.04026/) (2017)
- [Q-Align: Teaching LMMs for Visual Scoring via Discrete Text-Defined Levels](https://mlsanity.com/paper/2312.17090/) (2023)
- [Benchmarking Single Image Dehazing and Beyond](https://mlsanity.com/paper/1712.04143/) (2017)
- [Learning Deep CNN Denoiser Prior for Image Restoration](https://mlsanity.com/paper/1704.03264/) (2017)
- [R1-Onevision: Advancing Generalized Multimodal Reasoning through Cross-Modal Formalization](https://mlsanity.com/paper/2503.10615/) (2025)
- [DehazeNet: An End-to-End System for Single Image Haze Removal](https://mlsanity.com/paper/1601.07661/) (2016)
- [R1-VL: Learning to Reason with Multimodal Large Language Models via Step-wise Group Relative Policy Optimization](https://mlsanity.com/paper/2503.12937/) (2025)
- [Post-Training Quantization for Vision Transformer](https://mlsanity.com/paper/2106.14156/) (2021)

## Built on by (most cited first, 2 of 2)

- [Bridging Information Asymmetry: A Hierarchical Framework for Deterministic Blind Face Restoration](https://mlsanity.com/paper/2601.19506/) (2026)
- [RefReward-SR: LR-Conditioned Reward Modeling for Preference-Aligned Super-Resolution](https://mlsanity.com/paper/2603.24198/) (2026)

## Abstract (arXiv)

Post-training has become effective for high-level generation, but its role in low-level vision remains underexplored. Existing image restoration methods often rely on fixed pixel-wise fitting to ground-truth images, which can lead to over-smoothing and weak generalization. We propose IRPO, a GRPO-based post-training framework for deterministic restoration models. IRPO is built around two axes: data formulation and reward modeling. For data formulation, we select the 30% underperforming samples from the pre-training stage, which improves both accuracy and training efficiency. For reward modeling, we combine fidelity-oriented and quality-aware feedback with three components: a General Reward for structural fidelity, an Expert Reward that uses a Vision-Language Model as a coarse visual-quality judge, and a Restoration Reward for task-specific low-level cues. Experiments on six in-domain and five out-of-domain (OOD) benchmarks show that IRPO improves the AdaIR baseline by 0.93 dB on in-domain tasks and 3.43 dB on OOD settings. Our code can be shown in https://github.com/HaoxuanXU1024/IRPO.