Psychological AI First Aid for Mental Relaxation: A Privacy-Preserving, EmotionAware Support Framework

Author : Sandhya e k

Abstract : Dhiya AI is an intelligent and empathetic mental health support system designed using advanced artificial intelligence techniques to provide accessible, private, and real-time emotional assistance. At the core, a multimodal deep learning model combined with affective computing and Natural Language Processing (NLP) is implemented to interpret user emotions based on facial expressions, written inputs, and conversational turns. A mood analyzer is developed through ConvLSTM networks which can handle temporal patterns in the textual data to separate emotional fluctuations as well as long term mental rides while deep learning-based facial expression recognition used for affect detection. The outputs of two models are merged into a single emotional representation to perform accurate and individualized interventions. Furthermore, a Generative AI–based chatbot furnishes supportive and contextaware advice with respect to mental health principles. With generative AI and state-of-the-art emotion analysis, Dhiya is a wellrounded mental health assistant for users with zero registration requirements to become the go-to source for increasing emotional awareness, detecting early signs of distress all while providing empathy

Keywords : Natural Language Processing (NLP), Affective Computing, ConvLSTM, Facial Expression Recognition, Generative AI, Multimodal Emotion Analysis, Mental Health Support, Deep Learning.

Conference Name : International Conference on Data Science for Healthcare, Finance, and Social Sciences (ICDSHFSS-26)

Conference Place : Chennai, India

Conference Date : 10th Jan 2026