Vision-Language Models vs Human: Perceptual Image Quality Assessment

• URL: http://arxiv.org/abs/2603.24578v1
• Date: Wed, 25 Mar 2026 17:54:07 GMT
• Title: Vision-Language Models vs Human: Perceptual Image Quality Assessment
• Abstract summary: Psychophysical experiments remain the most reliable approach for perceptual image quality assessment (IQA)<n>We investigate whether Vision Language Models (VLMs) can approximate human perceptual judgments across three image quality scales.
• Score: 0.01116979912801043
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Psychophysical experiments remain the most reliable approach for perceptual image quality assessment (IQA), yet their cost and limited scalability encourage automated approaches. We investigate whether Vision Language Models (VLMs) can approximate human perceptual judgments across three image quality scales: contrast, colorfulness and overall preference. Six VLMs four proprietary and two openweight models are benchmarked against psychophysical data. This work presents a systematic benchmark of VLMs for perceptual IQA through comparison with human psychophysical data. The results reveal strong attribute dependent variability models with high human alignment for colorfulness (ρup to 0.93) underperform on contrast and vice-versa. Attribute weighting analysis further shows that most VLMs assign higher weights to colorfulness compared to contrast when evaluating overall preference similar to the psychophysical data. Intramodel consistency analysis reveals a counterintuitive tradeoff: the most self consistent models are not necessarily the most human aligned suggesting response variability reflects sensitivity to scene dependent perceptual cues. Furthermore, human-VLM agreement is increased with perceptual separability, indicating VLMs are more reliable when stimulus differences are clearly expressed.

Related papers

• Preference learning in shades of gray: Interpretable and bias-aware reward modeling for human preferences [8.680081568962999]
  This study investigates the limits of current approaches to learning human preferences in language models.<n>We propose a feature-augmented framework to better capture the multidimensional nature of human judgment.
  arXiv  Detail & Related papers  (2026-04-01T18:26:16Z)
• Do Machines Fail Like Humans? A Human-Centred Out-of-Distribution Spectrum for Mapping Error Alignment [1.974820485832244]
  We propose a human-centred framework that redefines the degree of OOD as a spectrum of human perceptual difficulty.<n>We apply this framework to object recognition and reveal unique, regime-dependent model-human alignment rankings and profiles across deep learning architectures.
  arXiv  Detail & Related papers  (2026-03-08T04:51:39Z)
• Visual Affect Analysis: Predicting Emotions of Image Viewers with Vision-Language Models [2.2023261946811563]
  Vision-language models (VLMs) show promise as tools for inferring affect from visual stimuli at scale.<n>We benchmarked nine VLMs, ranging from state-of-the-art proprietary models to open-source models, on three psycho-metrically validated affective image datasets.
  arXiv  Detail & Related papers  (2026-01-27T17:56:21Z)
• VLIC: Vision-Language Models As Perceptual Judges for Human-Aligned Image Compression [83.36460501519203]
  Vision-Language Models for Image Compression (VLIC) is a diffusion-based image compression system designed to be post-trained with binary VLM judgments.<n>We show that calibrating this system on VLM judgments produces competitive or state-of-the-art performance on human-aligned visual compression depending on the dataset.
  arXiv  Detail & Related papers  (2025-12-17T18:52:55Z)
• Building Reasonable Inference for Vision-Language Models in Blind Image Quality Assessment [7.969076042774561]
  We analyze the factors that cause contradictory assessments and instability.<n>We introduce a two-stage tuning method that explicitly separates visual perception from quality inference.
  arXiv  Detail & Related papers  (2025-12-10T11:50:42Z)
• RoboView-Bias: Benchmarking Visual Bias in Embodied Agents for Robotic Manipulation [67.38036090822982]
  We propose RoboView-Bias, the first benchmark specifically designed to quantify visual bias in robotic manipulation.<n>We create 2,127 task instances that enable robust measurement of biases induced by individual visual factors and their interactions.<n>Our results highlight that systematic analysis of visual bias is a prerequisite for developing safe and reliable general-purpose embodied agents.
  arXiv  Detail & Related papers  (2025-09-26T13:53:25Z)
• Contrast Sensitivity in Multimodal Large Language Models: A Psychophysics-Inspired Evaluation [37.9406446788251]
  We introduce a behavioural method for estimating the Contrast Sensitivity Function (CSF) in Multimodal Large Language Models (MLLMs)<n>Models are queried with structured prompts while viewing noise-based stimuli filtered at specific spatial frequencies.<n>Our results reveal that some models resemble human CSFs in shape or scale, but none capture both.
  arXiv  Detail & Related papers  (2025-08-14T06:07:44Z)
• Investigating VLM Hallucination from a Cognitive Psychology Perspective: A First Step Toward Interpretation with Intriguing Observations [60.63340688538124]
  Hallucination is a long-standing problem that has been actively investigated in Vision-Language Models (VLMs)<n>Existing research commonly attributes hallucinations to technical limitations or sycophancy bias, where the latter means the models tend to generate incorrect answers to align with user expectations.<n>In this work, we introduce a psychological taxonomy, categorizing VLMs' cognitive biases that lead to hallucinations, including sycophancy, logical inconsistency, and a newly identified VLMs behaviour: appeal to authority.
  arXiv  Detail & Related papers  (2025-07-03T19:03:16Z)
• VHELM: A Holistic Evaluation of Vision Language Models [75.88987277686914]
  We present the Holistic Evaluation of Vision Language Models (VHELM) VHELM aggregates various datasets to cover one or more of the 9 aspects: visual perception, knowledge, reasoning, bias, fairness, multilinguality, robustness, toxicity, and safety. Our framework is designed to be lightweight and automatic so that evaluation runs are cheap and fast.
  arXiv  Detail & Related papers  (2024-10-09T17:46:34Z)
• Mind the Uncertainty in Human Disagreement: Evaluating Discrepancies between Model Predictions and Human Responses in VQA [26.968874222330978]
  This study focuses on the Visual Question Answering (VQA) task. We evaluate how well vision-language models correlate with the distribution of human responses.
  arXiv  Detail & Related papers  (2024-09-17T13:44:25Z)
• Opinion-Unaware Blind Image Quality Assessment using Multi-Scale Deep Feature Statistics [54.08757792080732]
  We propose integrating deep features from pre-trained visual models with a statistical analysis model to achieve opinion-unaware BIQA (OU-BIQA) Our proposed model exhibits superior consistency with human visual perception compared to state-of-the-art BIQA models.
  arXiv  Detail & Related papers  (2024-05-29T06:09:34Z)
• When No-Reference Image Quality Models Meet MAP Estimation in Diffusion Latents [92.45867913876691]
  No-reference image quality assessment (NR-IQA) models can effectively quantify perceived image quality.<n>We show that NR-IQA models can be plugged into the maximum a posteriori (MAP) estimation framework for image enhancement.
  arXiv  Detail & Related papers  (2024-03-11T03:35:41Z)
• Perceptual Attacks of No-Reference Image Quality Models with Human-in-the-Loop [113.75573175709573]
  We make one of the first attempts to examine the perceptual robustness of NR-IQA models. We test one knowledge-driven and three data-driven NR-IQA methods under four full-reference IQA models. We find that all four NR-IQA models are vulnerable to the proposed perceptual attack.
  arXiv  Detail & Related papers  (2022-10-03T13:47:16Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.