Automatic Negative Thoughts using Multimodal Approach with Deep Neural Network

Automatic Negative Thoughts are negative beliefs and involuntary responses to situations based on core beliefs about oneself or others. If left untreated, automatic negative thoughts can lead to depression, self-doubt, anger and anxiety. In this study, we present a new approach to identify automatic negative thoughts by detecting emotion, thinking error and situation which will help to overcome complications like depression. We present a multi-modal approach to identify dysfunctional automatic thoughts through face, speech and text modalities using deep neural network. Proposed model achieved an accuracy of 65% for facial emotion, 80% for thinking error and 84 % for situation.