GRENZE International Journal of Engineering and Technology
Vol. 10
(2024), Issue 1
Augmented Humanoid Counselor for Mental Health using LSTM Neural Network
Authors
Sonali Sharma, Vinayak Unde, Sakshi Jadhav, Dev Prajapati, Deepshikha Chaturvedi
Abstract
In every period of life, from childhood and adolescence to maturity, mental health is vital. Good mental health allows individuals to enhance their overall quality of life. However, due to a lack of willingness, many people neither share their mental health issues with their close ones nor seek professional help. To address this issue, we have developed a lifelike counselor who can offer effective and compassionate mental health conversations by uniting natural language processing and augmented reality. The Augmented Humanoid Counselor, a virtual counselor equipped with a conversational model, is designed to give a human touch by assisting individuals to manage their mental health. We have designed various neural network-based conversational models using Long Short-Term Memory (LSTM), Bi-directional LSTM, and LSTM with an attention layer. The LSTM model, enhanced with an attention layer, enables itself to concentrate on the input sequence's most instructive segments while disregarding the less important ones, and thus has demonstrated its ability to provide more interactive, relatable responses compared to other models. This conversational model is used by our Augmented Humanoid Counselor to respond to the user in a sympathetic way to give them support for their mental health and make them feel better
Pages:
840 - 847