Test-Time Training for Depression Detection

arXiv:2404.05071v1 Announce Type: new
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 …

abstract arxiv cs.lg cs.sd datasets demographics depression detection distribution eess.as environment environments gender however noise practice recording test train training type