采用语音的抑郁症诊断研究述评

Review of Research on Depression Diagnosis with Speech Signal

采用语音声学特征作为客观指标对抑郁症进行诊断研究,是近年来新兴的交叉研究课题之一.本文介绍了近年来国外抑郁症语音研究的状况,阐述了现有各种抑郁症诊断的方法,总结了具有代表性的抑郁症语音数据库.并在此基础上归纳了基于语音的抑郁症诊断研究中常用的各类特征参数,对当前基于语音的抑郁症诊断研究领域所面临的挑战进行了展望,旨在尽可能全面地揭示抑郁症和语音特征的密切关系,为相关研究人员提供有价值的学术参考.

The depression diagnosis by using acoustic features as an objective predictor is a state-of-the-art cross-cover research. This paper firstly introduces various existing approaches of diagnosing depression by reviewing the speech-based depression diagnosis research in recent years. Then the paper summarizes the speech corpora and different acoustic features used for speech-based depression diagnosis. Finally the paper draws a conclusion of the challenges faced by current researches of speech-based depression diagnosis. The paper not only can reveal the relationship between depression and acoustic features,the studies on depression diagnosis are also can be promoted by providing the valuable academic references.