The project develops a platform (DTracker) that allows to collect user data using a smartphone, that can be analyzed automatically and by doctors to detect changes in mood and depressive states. The mobile app collects information of the patient activity and asks the patients to explain their feelings to the camera.  This information is sent to a central server that performs the analysis of the activity, the voice and the user face to guess the user mod, and raise the corresponding alarms when it deteriorates.

The TALN group has been involved in the development of the voice analysis module. The voice module performs the extraction of different features from the recordings. The solution provided adds prosody features to the most commonly used acoustic features.  A mood status model has been computed from this data.