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 2404.05071cs.LGcs.SDeess.AS
收藏

用于抑郁症检测的测试时训练

Test-Time Training for Depression Detection

Dalhousie University(达尔豪斯大学);Vector Institute(矢量研究所);Nova Scotia Health(新斯科舍省卫生机构)

Sri Harsha Dumpala, Chandramouli Shama Sastry, Rudolf Uher, Sageev Oore

AI总结:

本文针对抑郁症检测中训练与测试分布不一致导致性能下降的问题,引入测试时训练(TTT),显著提升模型在噪声、性别偏见及跨数据集场景下的鲁棒性。

摘要:

以往关于抑郁症检测的研究使用在相似环境中收集的数据集来训练和测试模型。然而在实际应用中,无法保证训练分布与测试分布完全相同。分布偏移可能由录音环境(例如背景噪声)和人口统计学特征(例如性别、年龄等)的变化所引入。这种分布偏移会出人意料地导致抑郁症检测模型的性能严重下降。本文分析了测试时训练(TTT)的应用,以提升为抑郁症检测所训练模型的鲁棒性。与模型的常规测试相比,我们发现 TTT 能够在多种分布偏移下显著提升模型的鲁棒性,这些偏移由以下因素引入:(a) 背景噪声,(b) 性别偏见,以及 (c) 数据收集与整理流程(即训练样本和测试样本来自不同的数据集)。

英文摘要:

Previous works on depression detection use datasets collected in similar environments to train and test the models. In practice, however, the train and test distributions cannot be guaranteed to be identical. Distribution shifts can be introduced due to variations such as recording environment (e.g., background noise) and demographics (e.g., gender, age, etc). Such distributional shifts can surprisingly lead to severe performance degradation of the depression detection models. In this paper, we analyze the application of test-time training (TTT) to improve robustness of models trained for depression detection. When compared to regular testing of the models, we find TTT can significantly improve the robustness of the model under a variety of distributional shifts introduced due to: (a) background-noise, (b) gender-bias, and (c) data collection and curation procedure (i.e., train and test samples are from separate datasets).

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。

↑