mlsanityData to 14 Oct 2026

UniPercept: Towards Unified Perceptual-Level Image Understanding across Aesthetics, Quality, Structure, and Texture

arXiv 2512.21675 ↗First posted 25 Dec 2025cs.CVcited by 22

On the map: Multimodal LLMs / vision-language models → MLLM training and visual instruction tuning

Abstract (arXiv)

Multimodal large language models (MLLMs) have achieved remarkable progress in visual understanding tasks such as visual grounding, segmentation, and captioning. However, their ability to perceive perceptual-level image features remains limited. In this work, we present UniPercept-Bench, a unified framework for perceptual-level image understanding across three key domains: Aesthetics, Quality, Structure and Texture. We establish a hierarchical definition system and construct large-scale datasets to evaluate perceptual-level image understanding. Based on this foundation, we develop a strong baseline UniPercept trained via Domain-Adaptive Pre-Training and Task-Aligned RL, enabling robust generalization across both Visual Rating (VR) and Visual Question Answering (VQA) tasks. UniPercept outperforms existing MLLMs on perceptual-level image understanding and can serve as a plug-and-play reward model for text-to-image generation. This work defines Perceptual-Level Image Understanding in the era of MLLMs and, through the introduction of a comprehensive benchmark together with a strong baseline, provides a solid foundation for advancing perceptual-level multimodal image understanding.

Key facts. Cited by 22 papers in the map (data through 14 Oct 2026). Placed in MLLM training and visual instruction tuning, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 37 of its 59 references that are in the map. First posted on arXiv 25 Dec 2025.

Builds on

37 of 59 references.

Closest ancestors · 2

Same direction, newest first

  1. 2025LLaVA-OneVision-1.5: Fully Open Framework for Democratized Multimodal Training28 Sep
  2. 2023MMBench: Is Your Multi-modal Model an All-around Player?12 Jul

Other references · 35

  1. 2023GPT-4 Technical Reportcited by 18,396
  2. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 9,379
  3. 2025Qwen2.5-VL Technical Reportcited by 5,587
  4. 2013Describing Textures in the Wildcited by 2,125
  5. 2023MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning Benchmark for Expert AGIcited by 2,080
  6. 2025InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Modelscited by 1,834
  7. 2023MathVista: Evaluating Mathematical Reasoning of Foundation Models in Visual Contextscited by 1,614
  8. 2017Enhanced Deep Residual Networks for Single Image Super-Resolutioncited by 1,552
  9. 2021MUSIQ: Multi-scale Image Quality Transformercited by 1,081
  10. 2023Pick-a-Pic: An Open Dataset of User Preferences for Text-to-Image Generationcited by 977
  11. 2024Are We on the Right Way for Evaluating Large Vision-Language Models?cited by 914
  12. 2025Emerging Properties in Unified Multimodal Pretrainingcited by 897
23 more
  1. 2024MMMU-Pro: A More Robust Multi-discipline Multimodal Understanding Benchmarkcited by 512
  2. 2023Q-Align: Teaching LMMs for Visual Scoring via Discrete Text-Defined Levelscited by 503
  3. 2023OCRBench: On the Hidden Mystery of OCR in Large Multimodal Modelscited by 476
  4. 2024ELLA: Equip Diffusion Models with LLM for Enhanced Semantic Alignmentcited by 475
  5. 2025GLM-4.5V and GLM-4.1V-Thinking: Towards Versatile Multimodal Reasoning with Scalable Reinforcement Learningcited by 466
  6. 2025BLIP3-o: A Family of Fully Open Unified Multimodal Models-Architecture, Training and Datasetcited by 376
  7. 2019KonIQ-10k: An ecologically valid database for deep learning of blind image quality assessmentcited by 355
  8. 2024SANA: Efficient High-Resolution Image Synthesis with Linear Diffusion Transformerscited by 340
  9. 2024Scaling Up to Excellence: Practicing Model Scaling for Photo-Realistic Image Restoration In the Wildcited by 286
  10. 2023Q-Bench: A Benchmark for General-Purpose Foundation Models on Low-level Visioncited by 249
  11. 2025HPSv3: Towards Wide-Spectrum Human Preference Scorecited by 211
  12. 2025SANA 1.5: Efficient Scaling of Training-Time and Inference-Time Compute in Linear Diffusion Transformercited by 165
  13. 2023Q-Instruct: Improving Low-level Visual Abilities for Multi-modality Foundation Modelscited by 152
  14. 2025Teaching Large Language Models to Regress Accurate Image Quality Scores using Score Distributioncited by 141
  15. 2023Depicting Beyond Scores: Advancing Image Quality Assessment through Multi-modal Language Modelscited by 105
  16. 2025Q-Insight: Understanding Image Quality via Visual Reinforcement Learningcited by 98
  17. 2024Towards Open-ended Visual Quality Comparisoncited by 85
  18. 2024AesBench: An Expert Benchmark for Multimodal Large Language Models on Image Aesthetics Perceptioncited by 65
  19. 2024UNIAA: A Unified Multi-modal Image Aesthetic Assessment Baseline and Benchmarkcited by 42
  20. 2025ArtiMuse: Fine-Grained Image Aesthetics Assessment with Joint Scoring and Expert-Level Understandingcited by 41
  21. 2018PIRM Challenge on Perceptual Image Enhancement on Smartphones: Reportcited by 38
  22. 2024Q-Ground: Image Quality Grounding with Large Multi-modality Modelscited by 30
  23. 2025Next Token Is Enough: Realistic Image Quality and Aesthetic Scoring with Multimodal Large Language Modelcited by 12

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 22.

22 papers.

Sort:

By area:

  1. 2026GEMS: Agent-Native Multimodal Generation with Memory and Skillscited by 21
  2. 2026PiD: Fast and High-Resolution Latent Decoding with Pixel Diffusioncited by 7
  3. 2026ERNIE-Image Technical Reportcited by 6
  4. 2026AIBench: Evaluating Visual-Logical Consistency in Academic Illustration Generationcited by 4
  5. 2026AesRM: Improving Video Aesthetics with Expert-Level Feedbackcited by 3
  6. 2026SANA-Video 2.0: Hybrid Linear Attention with Attention Residuals for Efficient Video Generationcited by 2
  7. 2026LucidNFT: LR-Anchored Multi-Reward Preference Optimization for Flow-Based Real-World Super-Resolutioncited by 1
  8. 2026One Model, Two Minds: Task-Conditioned Reasoning for Unified Image Quality and Aesthetic Assessmentcited by 1
  9. 2026Sketch Then Paint: Hierarchical Reinforcement Learning for Diffusion Multi-Modal Large Language Modelscited by 1
  10. 2026Hard Vision, Easy Vision: What GPT-6 Astra Reveals Across Computer Visioncited by 1
  11. 2026Accelerating Masked Image Generation by Learning Controlled Latent Dynamicscited by 0
  12. 2026DynT2I-Eval: A Dynamic Evaluation Framework for Text-to-Image Modelscited by 0
10 more
  1. 2026PhotoFlow: Agentic 3D Virtual Photography Missionscited by 0
  2. 2026HiTokSR: A Coarse-to-Fine Tokenizer with Hierarchical Codebooks for High-Fidelity Real-World Image Super-Resolutioncited by 0
  3. 2026ShotCrop3: Cropping Human-Centric Images into Cinematic Triple-Shot Compositionscited by 0
  4. 2026Beyond Absolute Scores: Relative Edit-induced Difference for Generalizable Image Aesthetic Assessmentcited by 0
  5. 2026COMPASS: Grounding Composition-Intent Guidance in Unified Multimodal Modelscited by 0
  6. 2026AesCanvas: A Large-Scale Dataset and Benchmark for Aesthetic Critique and Contextual Suitabilitycited by 0
  7. 2026DramaChain Bench: An End-to-End Benchmark for Short-Drama Generationcited by 0
  8. 2026SoL-Refiner: Speed-of-Light One-Step Refinement for High-Resolution Videocited by 0
  9. 2026PortraitAes: Intent-Conditioned Structured Portrait Aesthetic Assessmentcited by 0
  10. 2026Golden Anchor for Video Generationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025UniModel: A Visual-Only Framework for Unified Multimodal Understanding and Generation
  2. 2025Perceiving Beyond Language Priors: Enhancing Visual Comprehension and Attention in Multimodal Models
  3. 2025UniPixel: Unified Object Referring and Segmentation for Pixel-Level Visual Reasoning
  4. 2024Q-Bench+: A Benchmark for Multi-modal Foundation Models on Low-level Vision from Single Images to Pairs
  5. 2024JourneyBench: A Challenging One-Stop Vision-Language Understanding Benchmark of Generated Images
  6. 2025UniGen: Enhanced Training & Test-Time Strategies for Unified Multimodal Understanding and Generation
  7. 2025Growing Visual Generative Capacity for Pre-Trained MLLMs
  8. 2024UniQA: Unified Vision-Language Pre-training for Image Quality and Aesthetic Assessment