一句话：提出 B-AIS 自动化流程，用领域语义黑盒评估 AI 软件中的视觉感知，以支持安全关键部署前的模型智能判断。
AI 赋能软件系统广泛用于自动驾驶、空中与海上等视觉任务，部署到公共道路等安全关键环境前，评估模型智能至关重要。论文提出 B-AIS，一个面向领域语义的自动化黑盒评估流程，用于评估 AI 赋能软件中的视觉感知能力。其核心思想是以领域语义为参照，在无法访问模型内部信息的情况下自动执行评估，从而判断视觉感知是否符合任务领域要求。摘要未报告具体实验设置与量化结果，但强调该方法服务于安全关键部署前的人工评估需求。
AI-enabled software systems (AIS) are prevalent in a wide range of applications, such as visual tasks of autonomous systems, extensively deployed in automotive, aerial, and naval domains. Hence, it is crucial for humans to evaluate the model’s intelligence before AIS is deployed to safety-critical environments, such as public roads.
DOI 原文 ·
@inproceedings{paperbot7157,
title = {B-AIS: An Automated Process for Black-box Evaluation of Visual Perception in AI-enabled Software against Domain Semantics},
author = {Hamed Barzamini and Mona Rahimi},
booktitle = {ASE},
year = {2022},
doi = {10.1145/3551349.3561162}
}