Topic Modeling Based Multi-modal Depression Detection

Abstract

Major depressive disorder is a common mental disorder that affects almost 7% of the adult U.S. population. The 2017 Audio/Visual Emotion Challenge (AVEC) asks participants to build a model to predict depression levels based on the audio, video, and text of an interview ranging between 7-33 minutes. Since averaging features over the entire interview will lose most temporal information, how to discover, capture, and preserve useful temporal details for such a long interview are significant challenges. Therefore, we propose a novel topic modeling based approach to perform context-aware analysis of the recording. Our experiments show that the proposed approach outperforms context-unaware methods and the challenge baselines for all metrics.

arXiv commentProceedings of the 7th Audio/Visual Emotion Challenge and Workshop (AVEC). (Official Depression Challenge Winner)

References in corpus (1)

• AVEC 2016 - Depression, Mood, and Emotion Recognition Workshop and Challenge

Cited by in corpus (9)

• Measuring Depression Symptom Severity from Spoken Language and 3D Facial Expressions
• Self-supervised representations in speech-based depression detection
• Text-based depression detection on sparse data
• MoodCapture: Depression Detection Using In-the-Wild Smartphone Images
• HiQuE: Hierarchical Question Embedding Network for Multimodal Depression Detection
• Two-stage Temporal Modelling Framework for Video-based Depression Recognition using Graph Representation
• CANAMRF: An Attention-Based Model for Multimodal Depression Detection
• Vision based body gesture meta features for Affective Computing
• Multimodal Depression Severity Prediction from medical bio-markers using Machine Learning Tools and Technologies