一句话：针对 DNN 测试与优化中标注数据昂贵的问题，提出面向深度神经网络的自适应测试选择方法。
深度神经网络（DNN）在过去十年取得巨大进展，DNN 驱动的软件应用虽能完成多种任务，却也可能产生错误行为并造成重大损失。为发现错误行为并提升 DNN 应用质量，开发者通常需要大量带标签数据用于模型测试与优化；然而，从应用场景采集多样数据并进行正确标注往往代价高昂、耗时严重。论文题目指向“自适应测试选择”作为应对思路：通过自适应地选择测试样本或数据，更有效利用有限标注预算来暴露 DNN 缺陷并支持测试与优化。但给定摘要仅阐述上述动机与数据瓶颈，未披露具体选择准则、算法实现及实验评估结果。
Deep neural networks (DNN) have achieved tremendous development in the past decade. While many DNN-driven software applications have been deployed to solve various tasks, they could also produce incorrect behaviors and result in massive losses. To reveal the incorrect behaviors and improve the quality of DNN-driven applications, developers often need rich labeled data for the testing and optimization of DNN models. However, in practice, collecting diverse data from application scenarios and labeling them properly is often a highly expensive and time-consuming task.
DOI 原文 ·
@inproceedings{paperbot6910,
title = {Adaptive test selection for deep neural networks},
author = {Xinyu Gao and Yang Feng and Yining Yin and Zixi Liu and Zhenyu Chen and Baowen Xu},
booktitle = {ICSE},
year = {2022},
doi = {10.1145/3510003.3510232}
}