Task-Utility IQA: Rethinking Image Quality Assessment in Embodied Tasks

查看原文 / View Original 编辑 / Edit

Computer Science > Computer Vision and Pattern Recognition arXiv:2610.35905 (cs) [Submitted on 30 Sep 2026] Title: Task-Utility IQA: Rethinking Image Quality Assessment in Embodied Tasks Authors: Jirong Zha, Yanwei Liao, Yemin Wang, Xiangmin Yi, Siqi Peng, Yingfeng Chen, Chen Gao, Xinlei Chen.

计算机科学 > 计算机视觉与模式识别 arXiv:2610.35905 (cs) [提交于 2026 年 9 月 30 日] 标题：任务效用图像质量评估：重新思考具身任务中的图像质量评估作者：Jirong Zha, Yanwei Liao, Yemin Wang, Xiangmin Yi, Siqi Peng, Yingfeng Chen, Chen Gao, Xinlei Chen。

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.

摘要：图像质量评估 (IQA) 通过判断视觉质量是否满足下游任务来支撑具身成像流程，然而大多数现有方法学习的是与静态基准上通用人类评分相一致的任务无关分数。这一目标与具身和交互式环境不匹配，在这些环境中，图像的充分性取决于任务目标、上下文以及塑造智能体决策的行动需求。

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.

我们认为，IQA 应从分数回归转向由具身任务效用定义的以目标为条件的判断。为此，我们提出了一种基于多模态大模型 (MLLM) 的具身 IQA 智能体框架，该框架能够推理任务需求、建立质量证据基础，并生成面向决策的评估。

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.

一项紧凑的 VisDrone 证据研究表明，传统的感知指标和通用的基于 MLLM 的 IQA 分数对于下游视觉语言模型 (VLM) 任务的成功而言是较弱的代理指标，而以任务为条件的效用分数能更好地预测视觉输入是否支持目标任务。这些发现支持将 IQA 重新构建为面向任务效用的具身、可解释且具有决策意识的视觉感知。