Abstract
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).
Abstract (translated)
之前的抑郁症检测工作使用了类似于相同环境收集的数据来训练和测试模型。然而,在实践中,训练和测试分布不能保证完全相同。由于诸如记录环境(例如,背景噪音)和人口统计学(例如,性别、年龄等)的差异,分布可能会发生偏移。这些分布偏移可能会导致抑郁症检测模型的性能严重下降。在本文中,我们分析了将测试时间训练(TTT)应用于改善为抑郁症检测训练模型增加鲁棒性的应用。与对模型的常规测试相比,我们发现TTT可以在由于以下原因引入的各种分布偏移上显著提高模型的鲁棒性:(a)背景噪音,(b)性别偏见,和(c)数据收集和编辑过程(即训练和测试样本来自不同的数据集)。
URL
https://arxiv.org/abs/2404.05071