MA-DLE: Speech-based Automatic Depression Level Estimation via Memory Augmentation

arXiv:2606.11197v1 Announce Type: cross Abstract: Speech-based automatic estimation of depression levels is essential for enabling early detection and timely intervention, particularly in resource-constrained mental health settings. In recent years, deep learning has demonstrated impressive success across various domains, including affective...

Läs hela artikeln hos källan.