arXiv cs.AI· Jirong Zha, Yanwei Liao, Yemin Wang, Xiangmin Yi, Siqi Peng, Yingfeng Chen, Chen Gao, Xinlei Chen·· 4 小时前AI 评分35
Task-Utility IQA：重新思考具身任务中的图像质量评估
Task-Utility IQA: Rethinking Image Quality Assessment in Embodied Tasks
AI 导读
一项 arXiv 研究提出，具身场景下的图像质量评估（IQA）应从通用评分回归转向由具身任务效用定义的目标条件判断，并给出基于 MLLM 的具身 IQA 智能体框架，可推理任务需求、定位质量证据并输出面向决策的评估。VisDrone 证据实验显示，传统感知指标和通用 MLLM 评分对下游 VLM 任务成功的预测较弱，而任务条件化效用评分能更好预测视觉输入是否支持目标任务。
正文 · 原文
Abstract:Image quality assessment (IQA) underpins embodied imaging pipelines by judging whether visual quality satisfies downstream tasks, yet most existing methods learn task-agnostic scores aligned with generic human ratings on static benchmarks. This objective mismatches embodied and interactive settings, where image adequacy depends on task goals, context, and action requirements that shape an agent's decisions. We argue that IQA should shift from score regression to goal-conditioned judgment defined by embodied task utility. To this end, we propose an MLLM-based embodied IQA agent framework that reasons about task requirements, grounds quality evidence, and produces decision-facing assessments. A compact VisDrone evidence study shows that conventional perceptual metrics and generic MLLM-based IQA scores are weak proxies for downstream VLM task success, while a task-conditioned utility score better predicts whether visual input supports the target task. These findings support a task-utility-oriented reframing of IQA for embodied, explainable, and decision-aware visual perception.
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.35905 [cs.CV] |
| (or arXiv:2610.35905v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2610.35905 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Jirong Zha [view email]
[v1]
Wed, 30 Sep 2026 03:27:47 UTC (8,990 KB)
来源：arXiv cs.AI · arxiv.org