mlsanityData to 14 Oct 2026

Enhancing Blind Face Restoration through Online Reinforcement Learning

arXiv 2509.23339 ↗First posted 27 Sep 2025cs.CVcited by 0

On the map: Deep reinforcement learning methods → RL with diffusion and flow models

Abstract (arXiv)

Blind Face Restoration (BFR) encounters inherent challenges in exploring its large solution space, leading to common artifacts like missing details and identity ambiguity in the restored images. To tackle these challenges, we propose a Likelihood-Regularized Policy Optimization (LRPO) framework, the first to apply online reinforcement learning (RL) to the BFR task. LRPO leverages rewards from sampled candidates to refine the policy network, increasing the likelihood of high-quality outputs while improving restoration performance on low-quality inputs. However, directly applying RL to BFR creates incompatibility issues, producing restoration results that deviate significantly from the ground truth. To balance perceptual quality and fidelity, we propose three key strategies: 1) a composite reward function tailored for face restoration assessment, 2) ground-truth guided likelihood regularization, and 3) noise-level advantage assignment. Extensive experiments demonstrate that our proposed LRPO significantly improves the face restoration quality over baseline methods and achieves state-of-the-art performance.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in RL with diffusion and flow models, one of 15 directions in Deep reinforcement learning methods. Builds on 47 of its 54 references that are in the map. First posted on arXiv 27 Sep 2025.

Builds on

47 of 54 references.

References · 47

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  2. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  3. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  4. 2017Proximal Policy Optimization Algorithmscited by 15,215
  5. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,177
  6. 2020Denoising Diffusion Implicit Modelscited by 9,781
  7. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 9,379
  8. 2015Deep Unsupervised Learning using Nonequilibrium Thermodynamicscited by 7,058
  9. 2018A Style-Based Generator Architecture for Generative Adversarial Networkscited by 6,013
  10. 2023Adding Conditional Control to Text-to-Image Diffusion Modelscited by 5,503
  11. 2023SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesiscited by 4,356
  12. 2019Analyzing and Improving the Image Quality of StyleGANcited by 3,588
35 more
  1. 2017Progressive Growing of GANs for Improved Quality, Stability, and Variationcited by 2,988
  2. 2025DAPO: An Open-Source LLM Reinforcement Learning System at Scalecited by 2,810
  3. 2021CLIPScore: A Reference-free Evaluation Metric for Image Captioningcited by 2,424
  4. 2016Concrete Problems in AI Safetycited by 1,642
  5. 2021MUSIQ: Multi-scale Image Quality Transformercited by 1,081
  6. 2022Exploring CLIP for Assessing the Look and Feel of Imagescited by 991
  7. 2023Training Diffusion Models with Reinforcement Learningcited by 982
  8. 2023Diffusion Model Alignment Using Direct Preference Optimizationcited by 790
  9. 2025Flow-GRPO: Training Flow Matching Models via Online RLcited by 661
  10. 2017The Perception-Distortion Tradeoffcited by 587
  11. 2022MANIQA: Multi-dimension Attention Network for No-Reference Image Quality Assessmentcited by 522
  12. 2023DPOK: Reinforcement Learning for Fine-tuning Text-to-Image Diffusion Modelscited by 450
  13. 2025DanceGRPO: Unleashing GRPO on Visual Generationcited by 413
  14. 2020PULSE: Self-Supervised Photo Upsampling via Latent Space Exploration of Generative Modelscited by 358
  15. 2021Towards Real-World Blind Face Restoration with Generative Facial Priorcited by 349
  16. 2024One-Step Effective Diffusion Network for Real-World Image Super-Resolutioncited by 246
  17. 2022Towards Robust Blind Face Restoration with Codebook Lookup Transformercited by 231
  18. 2020SRFlow: Learning the Super-Resolution Space with Normalizing Flowcited by 209
  19. 2021GAN Prior Embedded Network for Blind Face Restoration in the Wildcited by 197
  20. 2017FSRNet: End-to-End Learning Face Super-Resolution with Facial Priorscited by 170
  21. 2020GLEAN: Generative Latent Bank for Large-Factor Image Super-Resolutioncited by 158
  22. 2022VQFR: Blind Face Restoration with Vector-Quantized Dictionary and Parallel Decodercited by 116
  23. 2020Progressive Semantic-Aware Style Transformation for Blind Face Restorationcited by 99
  24. 2022DifFace: Blind Face Restoration with Diffused Error Contractioncited by 97
  25. 2025TempFlow-GRPO: When Timing Matters for GRPO in Flow Modelscited by 97
  26. 2025SimpleAR: Pushing the Frontier of Autoregressive Visual Generation through Pretraining, SFT, and RLcited by 86
  27. 2022RestoreFormer: High-Quality Blind Face Restoration from Undegraded Key-Value Pairscited by 85
  28. 2023DR2: Diffusion-based Robust Degradation Remover for Blind Face Restorationcited by 79
  29. 2019Progressive Face Super-Resolution via Attention to Facial Landmarkcited by 73
  30. 2024FlowIE: Efficient Image Enhancement via Rectified Flowcited by 38
  31. 2023Towards Real-World Blind Face Restoration with Generative Diffusion Priorcited by 22
  32. 2025AR-GRPO: Training Autoregressive Image Generation Models via Reinforcement Learningcited by 22
  33. 2025InterLCM: Low-Quality Images as Intermediate States of Latent Consistency Models for Effective Blind Face Restorationcited by 9
  34. 2025DiffusionReward: Enhancing Blind Face Restoration through Reward Feedback Learningcited by 5
  35. 2021VSpSR: Explorable Super-Resolution via Variational Sparse Representationcited by 4

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Unlocking the Potential of Diffusion Priors in Blind Face Restoration
  2. 2026Bridging Information Asymmetry: A Hierarchical Framework for Deterministic Blind Face Restoration
  3. 2024Towards Unsupervised Blind Face Restoration using Diffusion Prior
  4. 2024Visual Style Prompt Learning Using Diffusion Models for Blind Face Restoration
  5. 2024DR-BFR: Degradation Representation with Diffusion Models for Blind Face Restoration
  6. 2024CLR-Face: Conditional Latent Refinement for Blind Face Restoration Using Score-Based Diffusion Models
  7. 2025TD-BFR: Truncated Diffusion Model for Efficient Blind Face Restoration
  8. 2024AuthFace: Towards Authentic Blind Face Restoration with Face-oriented Generative Diffusion Prior