ME-IQA: Memory-Enhanced Image Quality Assessment via Re-Ranking
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918401230766080 |
|---|---|
| author | Fan, Kanglong Wu, Tianhe Wen, Wen Liu, Jianzhao Yang, Le Zhang, Yabin Liao, Yiting Li, Junlin Zhang, Li |
| author_facet | Fan, Kanglong Wu, Tianhe Wen, Wen Liu, Jianzhao Yang, Le Zhang, Yabin Liao, Yiting Li, Junlin Zhang, Li |
| contents | Reasoning-induced vision-language models (VLMs) advance image quality assessment (IQA) with textual reasoning, yet their scalar scores often lack sensitivity and collapse to a few values, so-called discrete collapse. We introduce ME-IQA, a plug-and-play, test-time memory-enhanced re-ranking framework. It (i) builds a memory bank and retrieves semantically and perceptually aligned neighbors using reasoning summaries, (ii) reframes the VLM as a probabilistic comparator to obtain pairwise preference probabilities and fuse this ordinal evidence with the initial score under Thurstone's Case V model, and (iii) performs gated reflection and consolidates memory to improve future decisions. This yields denser, distortion-sensitive predictions and mitigates discrete collapse. Experiments across multiple IQA benchmarks show consistent gains over strong reasoning-induced VLM baselines, existing non-reasoning IQA methods, and test-time scaling alternatives. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_20785 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | ME-IQA: Memory-Enhanced Image Quality Assessment via Re-Ranking Fan, Kanglong Wu, Tianhe Wen, Wen Liu, Jianzhao Yang, Le Zhang, Yabin Liao, Yiting Li, Junlin Zhang, Li Computer Vision and Pattern Recognition Reasoning-induced vision-language models (VLMs) advance image quality assessment (IQA) with textual reasoning, yet their scalar scores often lack sensitivity and collapse to a few values, so-called discrete collapse. We introduce ME-IQA, a plug-and-play, test-time memory-enhanced re-ranking framework. It (i) builds a memory bank and retrieves semantically and perceptually aligned neighbors using reasoning summaries, (ii) reframes the VLM as a probabilistic comparator to obtain pairwise preference probabilities and fuse this ordinal evidence with the initial score under Thurstone's Case V model, and (iii) performs gated reflection and consolidates memory to improve future decisions. This yields denser, distortion-sensitive predictions and mitigates discrete collapse. Experiments across multiple IQA benchmarks show consistent gains over strong reasoning-induced VLM baselines, existing non-reasoning IQA methods, and test-time scaling alternatives. |
| title | ME-IQA: Memory-Enhanced Image Quality Assessment via Re-Ranking |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2603.20785 |