mlsanityData to 14 Oct 2026

VisualQuality-R1: Reasoning-Induced Image Quality Assessment via Reinforcement Learning to Rank

arXiv 2505.14460 ↗First posted 20 May 2025cs.CVcited by 79

On the map: LLM reasoning → Multimodal reasoning

Abstract (arXiv)

DeepSeek-R1 has demonstrated remarkable effectiveness in incentivizing reasoning and generalization capabilities of large language models (LLMs) through reinforcement learning. Nevertheless, the potential of reasoning-induced computation has not been thoroughly explored in the context of image quality assessment (IQA), a task depending critically on visual reasoning. In this paper, we introduce VisualQuality-R1, a reasoning-induced no-reference IQA (NR-IQA) model, and we train it with reinforcement learning to rank, a learning algorithm tailored to the intrinsically relative nature of visual quality. Specifically, for a pair of images, we employ group relative policy optimization to generate multiple quality scores for each image. These estimates are used to compute comparative probabilities of one image having higher quality than the other under the Thurstone model. Rewards for each quality estimate are defined using continuous fidelity measures rather than discretized binary labels. Extensive experiments show that the proposed VisualQuality-R1 consistently outperforms discriminative deep learning-based NR-IQA models as well as a recent reasoning-induced quality regression method. Moreover, VisualQuality-R1 is capable of generating contextually rich, human-aligned quality descriptions, and supports multi-dataset training without requiring perceptual scale realignment. These features make VisualQuality-R1 especially well-suited for reliably measuring progress in a wide range of image processing tasks like super-resolution and image generation.

Key facts. Cited by 79 papers in the map (data through 14 Oct 2026). Placed in Multimodal reasoning, one of 17 directions in LLM reasoning. Builds on 34 of its 62 references that are in the map. First posted on arXiv 20 May 2025.

Builds on

34 of 62 references.

References · 34

  1. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 9,379
  2. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 8,065
  3. 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,330
  4. 2025Qwen2.5-VL Technical Reportcited by 5,587
  5. 2024GPT-4o System Cardcited by 4,588
  6. 2021SwinIR: Image Restoration Using Swin Transformercited by 1,887
  7. 2021MUSIQ: Multi-scale Image Quality Transformercited by 1,081
  8. 2022MANIQA: Multi-dimension Attention Network for No-Reference Image Quality Assessmentcited by 522
  9. 2023Q-Align: Teaching LMMs for Visual Scoring via Discrete Text-Defined Levelscited by 503
  10. 2017NIMA: Neural Image Assessmentcited by 389
  11. 2019KonIQ-10k: An ecologically valid database for deep learning of blind image quality assessmentcited by 355
  12. 2023DreamSim: Learning New Dimensions of Human Visual Similarity using Synthetic Datacited by 337
22 more
  1. 2023Blind Image Quality Assessment via Vision-Language Correspondence: A Multitask Learning Perspectivecited by 280
  2. 2016Deep Neural Networks for No-Reference and Full-Reference Image Quality Assessmentcited by 225
  3. 2015Massive Online Crowdsourced Study of Subjective and Objective Picture Qualitycited by 209
  4. 2020Fast Differentiable Sorting and Rankingcited by 197
  5. 2023Q-Instruct: Improving Low-level Visual Abilities for Multi-modality Foundation Modelscited by 152
  6. 2017RankIQA: Learning from Rankings for No-reference Image Quality Assessmentcited by 151
  7. 2023AGIQA-3K: An Open Database for AI-Generated Image Quality Assessmentcited by 142
  8. 2025Teaching Large Language Models to Regress Accurate Image Quality Scores using Score Distributioncited by 141
  9. 2020Uncertainty-Aware Blind Image Quality Assessment in the Laboratory and Wildcited by 134
  10. 2023Depicting Beyond Scores: Advancing Image Quality Assessment through Multi-modal Language Modelscited by 105
  11. 2025Q-Insight: Understanding Image Quality via Visual Reinforcement Learningcited by 98
  12. 2019dipIQ: Blind Image Quality Assessment by Learning-to-Rank Discriminable Image Pairscited by 73
  13. 2016Perceptual Quality Prediction on Authentically Distorted Images Using a Bag of Features Approachcited by 71
  14. 2024Adaptive Image Quality Assessment via Teaching Large Multimodal Model to Comparecited by 66
  15. 2024A Comprehensive Study of Multimodal Large Language Models for Image Quality Assessmentcited by 43
  16. 2021Continual Learning for Blind Image Quality Assessmentcited by 36
  17. 2025Toward Generalized Image Quality Assessment: Relaxing the Perfect Reference Quality Assumptioncited by 35
  18. 2013Learning to Rank for Blind Image Quality Assessmentcited by 25
  19. 2020Active Fine-Tuning from gMAD Examples Improves Blind Image Quality Assessmentcited by 24
  20. 2023Assessor360: Multi-sequence Network for Blind Omnidirectional Image Quality Assessmentcited by 17
  21. 2021Troubleshooting Blind Image Quality Models in the Wildcited by 14
  22. 2020Consolidated Dataset and Metrics for High-Dynamic-Range Image Qualitycited by 10

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 23; 2026: 56.

79 papers.

Sort:

By area:

  1. 2025NoHumansRequired: Autonomous High-Quality Image Editing Triplet Miningcited by 33
  2. 2025JarvisEvo: Towards a Self-Evolving Photo Editing Agent with Synergistic Editor-Evaluator Optimizationcited by 31
  3. 2026Diversity-Preserved Distribution Matching Distillation for Fast Visual Synthesiscited by 20
  4. 2025Generative Universal Verifier as Multimodal Meta-Reasonercited by 17
  5. 2026TextPecker: Rewarding Structural Anomaly Quantification for Enhancing Visual Text Renderingcited by 15
  6. 2026GEditBench v2: A Human-Aligned Benchmark for General Image Editingcited by 14
  7. 2026Reward Modeling from Natural Language Human Feedbackcited by 11
  8. 2025VQAThinker: Exploring Generalizable and Explainable Video Quality Assessment via Reinforcement Learningcited by 10
  9. 2025Self-Evolving Vision-Language Models for Image Quality Assessment via Voting and Rankingcited by 10
  10. 2025Reinforcement Learning for Large Model: A Surveycited by 9
  11. 2025PhotoFramer: Multi-modal Image Composition Instructioncited by 9
  12. 2026Rethinking Token-Level Policy Optimization for Multimodal Chain-of-Thoughtcited by 9
38 more
  1. 2026Leveraging Verifier-Based Reinforcement Learning in Image Editingcited by 9
  2. 2025Fine-grained Image Quality Assessment for Perceptual Image Restorationcited by 8
  3. 2025PocketSR: The Super-Resolution Expert in Your Pocket Mobilescited by 8
  4. 2025Unlocking the Essence of Beauty: Advanced Aesthetic Reasoning with Relative-Absolute Policy Optimizationcited by 7
  5. 2026PiD: Fast and High-Resolution Latent Decoding with Pixel Diffusioncited by 7
  6. 2025PosterCopilot: Toward Layout Reasoning and Controllable Editing for Professional Graphic Designcited by 6
  7. 2026Beyond VLM-Based Rewards: Diffusion-Native Latent Reward Modelingcited by 6
  8. 2026CoCo: Code as CoT for Text-to-Image Preview and Rare Concept Generationcited by 6
  9. 2025UARE: A Unified Vision-Language Model for Image Quality Assessment, Restoration, and Enhancementcited by 5
  10. 2026CaC: Advancing Video Reward Models via Hierarchical Spatiotemporal Concentratingcited by 5
  11. 2026TEXTS-Diff: TEXTS-Aware Diffusion Model for Real-World Text Image Super-Resolutioncited by 4
  12. 2026GDPO-SR: Group Direct Preference Optimization for One-Step Generative Image Super-Resolutioncited by 4
  13. 2026OPERA: An Agent for Image Restoration with End-to-End Joint Planning-Execution Optimizationcited by 4
  14. 2026Thinking with Frames: Generative Video Distortion Evaluation via Frame Reward Modelcited by 3
  15. 2026CoVEBench: Can Video Editing Models Handle Complex Instructions?cited by 3
  16. 2025DP2O-SR: Direct Perceptual Preference Optimization for Real-World Image Super-Resolutioncited by 2
  17. 2025Test-Time Preference Optimization for Image Restorationcited by 2
  18. 2025Beyond the Ground Truth: Enhanced Supervision for Image Restorationcited by 2
  19. 2025Plug In, Grade Right: Psychology-Inspired AGIQAcited by 2
  20. 2026From Global to Granular: Revealing IQA Model Performance via Correlation Surfacecited by 2
  21. 2026How far have we gone in Generative Image Restoration? A study on its capability, limitations and evaluation practicescited by 2
  22. 2026ColorFLUX: A Structure-Color Decoupling Framework for Old Photo Colorizationcited by 2
  23. 2025Exploring Instruction Data Quality for Explainable Image Quality Assessmentcited by 1
  24. 2025OmniQuality-R: Advancing Reward Models Through All-Encompassing Quality Assessmentcited by 1
  25. 2026Q-Probe: Scaling Image Quality Assessment to High Resolution via Context-Aware Agentic Probingcited by 1
  26. 2026Q-Hawkeye: Reliable Visual Policy Optimization for Image Quality Assessmentcited by 1
  27. 2026ArtifactLens: Hundreds of Labels Are Enough for Artifact Detection with VLMscited by 1
  28. 2026Seeing Beyond 8bits: Subjective and Objective Quality Assessment of HDR-UGC Videoscited by 1
  29. 2026LucidNFT: LR-Anchored Multi-Reward Preference Optimization for Flow-Based Real-World Super-Resolutioncited by 1
  30. 2026MDS-VQA: Model-Informed Data Selection for Video Quality Assessmentcited by 1
  31. 2026OARS: Process-Aware Online Alignment for Generative Real-World Image Super-Resolutioncited by 1
  32. 2026One Model, Two Minds: Task-Conditioned Reasoning for Unified Image Quality and Aesthetic Assessmentcited by 1
  33. 2026Banana100: Breaking NR-IQA Metrics by 100 Iterative Image Replications with Nano Banana Procited by 1
  34. 2026Multi-Granularity Reasoning for Image Quality Assessment via Attribute-Aware Reinforcement Learning to Rankcited by 1
  35. 2026ReasonEdit: Towards Interpretable Image Editing Evaluation via Reinforcement Learningcited by 1
  36. 2026Tool-IQA: Augmenting Image Quality Assessment with Simple Toolscited by 1
  37. 2026Beyond Time Shifts: Adapting Omni-LLM as a Reference-Free Evaluator for Generative Audio-Visual Modelscited by 1
  38. 2026IQA-T1: Tool-based Visual Evidence Reasoning for Image Quality Assessmentcited by 1

50 most cited of 79 All 79 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2025Reasoning as Representation: Rethinking Visual Reinforcement Learning in Image Quality Assessment
  2. 2025Q-Ponder: A Unified Training Pipeline for Reasoning-based Visual Quality Assessment
  3. 2026Zoom-IQA: Image Quality Assessment with Reliable Region-Aware Reasoning
  4. 2025Guiding Perception-Reasoning Closer to Human in Blind Image Quality Assessment
  5. 2025Building Reasonable Inference for Vision-Language Models in Blind Image Quality Assessment
  6. 2025VQ-Insight: Teaching VLMs for AI-Generated Video Quality Understanding via Progressive Visual Reinforcement Learning
  7. 2025Reason-RFT: Reinforcement Fine-Tuning for Visual Reasoning of Vision Language Models
  8. 2026RL-ScanIQA: Reinforcement-Learned Scanpaths for Blind 360degImage Quality Assessment