Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{194817,
        author = {Vedant Deogadkar and Ayush Kalaskar and Sahil Waghmode and Shriniwas Gavhane and Govind Pole},
        title = {MindMate: A Personalised AI-Based Platform for Mental Health Support},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {10},
        pages = {8129-8135},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=194817},
        abstract = {Mental health disorders, including anxiety, depression, and chronic stress, have become a significant global concern, affecting individuals across all age groups and socio-economic backgrounds. The growing demand for accessible, affordable, and personalized mental health support has led to the emergence of artificial intelligence (AI)-driven solutions. This paper presents a comprehensive survey of recent advancements in AI-based mental health systems and introduces MindMate, a personalized, privacy-aware, multimodal AI platform designed to enhance emotional well-being.
Unlike traditional approaches that rely on a single source of data, MindMate integrates multiple modalities, including textual journaling, selfie-based facial analysis, and short video-based emotional assessment, to provide a holistic understanding of a user’s mental state. The system performs sentiment analysis on journal entries, facial expression recognition from selfies, and behavioral and emotional cue extraction from video recordings. These inputs are processed using machine learning models and combined through a weighted multimodal fusion mechanism to generate accurate mood scores and meaningful psychological insights.
To ensure user trust and data security, the platform adopts a privacy-first architecture, where sensitive data such as images and videos are processed on-device using frameworks like TensorFlow.js and are not stored permanently. Only derived features such as emotion labels and mood scores are retained. Additionally, the system includes adaptive AI-driven interventions, such as personalized task recommendations, and early warning mechanisms for critical mental health conditions. In high-risk scenarios, the platform can trigger alerts to emergency contacts with user consent.
This survey emphasizes the effectiveness of multimodal AI in improving the accuracy, reliability, and personalization of mental health assessment systems. It also discusses current challenges, including model bias, data privacy concerns, and user engagement, while highlighting future research directions in intelligent mental health care systems.},
        keywords = {Artificial Intelligence, Emotion Detection, Mental Health, Multimodal Fusion, Natural Language Processing (NLP), On-Device Machine Learning, Privacy-Preserving Systems, Sentiment Analysis, TensorFlow.js},
        month = {March},
        }