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{206753,
        author = {Aparna N and Suresha D and Spoorthi B},
        title = {Multimodal Mental Health Detection using BERT and Audio Analysis},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {no},
        pages = {319-323},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=206753},
        abstract = {There have been cases of the increasing number of mental illnesses like depression and suicidal thoughts, and there is an urgent need for early diagnosis and treatment. In this research, the researchers have developed a unique multimodal system to determine one's mental status based on text analysis and voice recognition. The text recognition process has been done using fine-tuning of a BERT model, while Mel Frequency Cepstral Coefficients (MFCC) have been used to process the voice input. As for the results, the proposed system identifies the user input into one of the three categories, including Normal, Depressed, and High Suicide Risk, along with a corresponding confidence score and the risk value assigned to each input. The text-based model yields 94% accuracy, while the voice model attains 95% accuracy. As it can be seen, the integration of both approaches helps get better results, as linguistic context and emotions can be used for further analysis. It is also worth mentioning that the system has been implemented in real-time using the Streamlit web application.},
        keywords = {Mental Health Detection, BERT, Multimodal Learning, Emotion Recognition, Natural Language Processing, Suicide Risk Prediction.},
        month = {July},
        }