Reasoning as Representation: Rethinking Visual Reinforcement Learning in Image Quality Assessment
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911480730877952 |
|---|---|
| author | Zhao, Shijie Zhang, Xuanyu Li, Weiqi Li, Junlin Zhang, Li Xue, Tianfan Zhang, Jian |
| author_facet | Zhao, Shijie Zhang, Xuanyu Li, Weiqi Li, Junlin Zhang, Li Xue, Tianfan Zhang, Jian |
| contents | Reasoning-based image quality assessment (IQA) models trained through reinforcement learning (RL) exhibit exceptional generalization, yet the underlying mechanisms and critical factors driving this capability remain underexplored in current research. Moreover, despite their superior performance, these models incur inference energy usage and latency orders of magnitude higher than their earlier counterparts, restricting their deployment in specific scenarios. Through extensive experiments, this paper verifies and elaborates that through RL training, MLLMs leverage their reasoning capability to convert redundant visual representations into compact, cross-domain aligned text representations. This conversion is precisely the source of the generalization exhibited by these reasoning-based IQA models. Building on this fundamental insight, we propose a novel algorithm, RALI, which employs contrastive learning to directly align images with these generalizable text representations learned by RL. This approach eliminates the reliance on reasoning processes and even obviates the need to load an LLM. For the quality scoring task, this framework achieves generalization performance comparable to reasoning-based models while requiring less than 5% of their model parameters and inference time. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_11369 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Reasoning as Representation: Rethinking Visual Reinforcement Learning in Image Quality Assessment Zhao, Shijie Zhang, Xuanyu Li, Weiqi Li, Junlin Zhang, Li Xue, Tianfan Zhang, Jian Computer Vision and Pattern Recognition Reasoning-based image quality assessment (IQA) models trained through reinforcement learning (RL) exhibit exceptional generalization, yet the underlying mechanisms and critical factors driving this capability remain underexplored in current research. Moreover, despite their superior performance, these models incur inference energy usage and latency orders of magnitude higher than their earlier counterparts, restricting their deployment in specific scenarios. Through extensive experiments, this paper verifies and elaborates that through RL training, MLLMs leverage their reasoning capability to convert redundant visual representations into compact, cross-domain aligned text representations. This conversion is precisely the source of the generalization exhibited by these reasoning-based IQA models. Building on this fundamental insight, we propose a novel algorithm, RALI, which employs contrastive learning to directly align images with these generalizable text representations learned by RL. This approach eliminates the reliance on reasoning processes and even obviates the need to load an LLM. For the quality scoring task, this framework achieves generalization performance comparable to reasoning-based models while requiring less than 5% of their model parameters and inference time. |
| title | Reasoning as Representation: Rethinking Visual Reinforcement Learning in Image Quality Assessment |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2510.11369 |