arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2404.18316cs.CVcs.AIcs.LG
收藏

立场:在没有上下文的情况下不要解释视觉模型

Position: Do Not Explain Vision Models Without Context

Faculty of Mathematics and Information Science, Warsaw University of Technology(华沙理工大学数学与信息科学学院);Faculty of Mathematics, Informatics and Mechanics, University of Warsaw(华沙大学数学、信息学与力学学院)

Paulina Tomaszewska, Przemysław Biecek

AI总结:

本文指出当前视觉模型解释方法缺乏上下文信息,通过回顾流行方法、展示XAI失败案例和现实用例,提出新研究方向并主张解释从“哪里”转向“如何”。

摘要:

图片中的听诊器是否使相邻的人成为医生还是患者?这当然取决于两个对象的上下文关系。如果这很明显,为什么视觉模型的解释方法不使用上下文信息?在本文中,我们(1)通过指出最流行的计算机视觉模型解释方法未考虑上下文信息来进行回顾,(2)展示流行XAI方法的失败示例,(3)提供空间上下文发挥重要作用的现实用例示例,(4)提出可能更好地利用上下文信息来解释计算机视觉模型的新研究方向,(5)论证解释方法需要从“哪里”转变为“如何”。

英文摘要:

Does the stethoscope in the picture make the adjacent person a doctor or a patient? This, of course, depends on the contextual relationship of the two objects. If it's obvious, why don't explanation methods for vision models use contextual information? In this paper, we (1) review the most popular methods of explaining computer vision models by pointing out that they do not take into account context information, (2) show examples of failures of popular XAI methods, (3) provide examples of real-world use cases where spatial context plays a significant role, (4) propose new research directions that may lead to better use of context information in explaining computer vision models, (5) argue that a change in approach to explanations is needed from 'where' to 'how'.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑