Predicting Presence and Severity of Depression from Speech Using Emotional Transfer Learning

Abstract

A presentation of my internship project while at Hoffman-La Roche on using transfer learning from emotional speech to detect depression. We trained a Mixture of Experts consisting of gradient-boosted decision tree classifiers to classify happiness and sadness in datasets of acted emotional speech in English and German. The model was applied to a dataset of interviews with Danish speaking patients with first episode depression and matched healthy controls. We observed significant seperation between the two groups, and found patients in remission to speak similarly to controls. Further, we conducted experiments on the effect of removing background noise and speaker diarization, which showed consistent levels of background noise to be crucial for consistent inferences.

Date
Dec 16, 2020 14:00 — 15:00
Event
Data Science Hour
Location
Hoffman-La Roche
Basel, 94305
Click on the Slides button above to view the slides.

Blog post to follow.. (maybe)

Lasse Hansen
Lasse Hansen
PhD Student in Machine Learning for Healthcare

I am a PhD student at the Department of Clinical Medicine at Aarhus University studying how to use machine learning to improve patient outcomes in psychiatry. In particular, my focus is on applying methods from Natural Language Processing to electronic health records for early prediction of schizophrenia and bipolar disorder.