Depression Detection Using Stereo Crossview Global Attention with Fully Convolutional Neural Networks
Received: 13 June 2025 | Revised: 12 August 2025 | Accepted: 25 August 2025 | Online: 5 December 2025
Corresponding author: Kamalamma Puttaraju Impana
Abstract
Detection of depression involves identifying physiological, behavioral, and textual cues to recognize signs of mental distress. Patterns in text, speech, physiological signals, and facial expressions are employed to assess a person's emotional state. However, traditional methods often fail to interpret data patterns due to limited long-range dependencies, which minimize model performance in depression detection. This research proposes Stereo Cross-view Global Attention with Fully Convolutional Neural Network (SCGA-FCNN) to detect depression effectively. In traditional FCNN, SCGA is incorporated to capture global contextual dependencies among multiple views, enhancing spatial awareness and cross-view interactions, allowing the model to detect subtle emotions more efficiently. Support Vector Machine-Synthetic Minority Over-sampling Technique (SVM-SMOTE) is applied to balance data by generating synthetic minority class samples near the support vectors, which helps to preserve and define more accurate class boundaries. Compared to existing methods such as Embeddings from Language Models (ELMo) with SVM, the proposed SCGA-FCNN achieves a higher accuracy of 89.98%, 85.75%, and 94.17% for NLSAA, PHQ-9, and Dreaddit datasets, respectively.
Keywords:
depression, embeddings from language models, fully convolutional neural network, stereo cross-view global attention, support vector machineDownloads
References
B. H. Bhavani, M. Sreenatha, and N. C. Kundur, ''Diagnosis and Classification of Depressive Disorders using ML and DL Models,'' Engineering, Technology & Applied Science Research, vol. 15, no. 2, pp. 21383–21389, Apr. 2025. DOI: https://doi.org/10.48084/etasr.10017
M. K. Myee, R. D. C. Rebekah, T. Deepa, G. D. Zion, and K. Lokesh, ''Detection of Depression in Social Media Posts using Emotional Intensity Analysis,'' Engineering, Technology & Applied Science Research, vol. 14, no. 5, pp. 16207–16211, Oct. 2024. DOI: https://doi.org/10.48084/etasr.7461
J. P. Thekkekara, S. Yongchareon, and V. Liesaputra, ''An attention-based CNN-BiLSTM model for depression detection on social media text,'' Expert Systems with Applications, vol. 249, Sept. 2024, Art. no. 123834. DOI: https://doi.org/10.1016/j.eswa.2024.123834
A. Gopalakrishnan et al., ''Prenatal depression level prediction using ensemble based deep learning model,'' International Journal of Cognitive Computing in Engineering, vol. 6, pp. 267–279, Dec. 2025. DOI: https://doi.org/10.1016/j.ijcce.2024.12.002
Z. N. Vasha, B. Sharma, I. J. Esha, J. A. Nahian, and J. A. Polin, ''Depression detection in social media comments data using machine learning algorithms,'' Bulletin of Electrical Engineering and Informatics, vol. 12, no. 2, pp. 987–996, Apr. 2023. DOI: https://doi.org/10.11591/eei.v12i2.4182
E. Bao, A. Pérez, and J. Parapar, ''Explainable depression symptom detection in social media,'' Health Information Science and Systems, vol. 12, no. 1, Sept. 2024, Art. no. 47. DOI: https://doi.org/10.1007/s13755-024-00303-9
C. Xin and L. Q. Zakaria, ''Integrating Bert With CNN and BiLSTM for Explainable Detection of Depression in Social Media Contents,'' IEEE Access, vol. 12, pp. 161203–161212, 2024. DOI: https://doi.org/10.1109/ACCESS.2024.3488081
E. Kerz, S. Zanwar, Y. Qiao, and D. Wiechmann, ''Toward explainable AI (XAI) for mental health detection based on language behavior,'' Frontiers in Psychiatry, vol. 14, Dec. 2023, Art. no. 1219479. DOI: https://doi.org/10.3389/fpsyt.2023.1219479
S. Inamdar, R. Chapekar, S. Gite, and B. Pradhan, ''Machine Learning Driven Mental Stress Detection on Reddit Posts Using Natural Language Processing,'' Human-Centric Intelligent Systems, vol. 3, no. 2, pp. 80–91, June 2023. DOI: https://doi.org/10.1007/s44230-023-00020-8
N. Oryngozha, P. Shamoi, and A. Igali, ''Detection and Analysis of Stress-Related Posts in Reddit’s Acamedic Communities,'' IEEE Access, vol. 12, pp. 14932–14948, 2024. DOI: https://doi.org/10.1109/ACCESS.2024.3357662
A. Zafar, D. Aftab, R. Qureshi, Y. Wang, and H. Yan, “Multi-Explainable TemporalNet: An Interpretable Multimodal Approach using Temporal Convolutional Network for User-level Depression Detection,” in 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), Seattle, WA, USA, Jun. 2024, pp. 2258–2265. DOI: https://doi.org/10.1109/CVPRW63382.2024.00231
M. R. Febriansyah, Nicholas, R. Yunanda, and D. Suhartono, ''Stress detection system for social media users,'' Procedia Computer Science, vol. 216, pp. 672–681, Jan. 2023. DOI: https://doi.org/10.1016/j.procs.2022.12.183
L. Ilias, S. Mouzakitis, and D. Askounis, ''Calibration of Transformer-Based Models for Identifying Stress and Depression in Social Media,'' IEEE Transactions on Computational Social Systems, vol. 11, no. 2, pp. 1979–1990, Apr. 2024. DOI: https://doi.org/10.1109/TCSS.2023.3283009
''PHQ-9 Depression Assessment.'' Available: https://www.kaggle.com/datasets/thedevastator/phq-9-depression-assessment.
Y. Yan and Q. Huang ''Dataset of depression and anxiety among the elderly derived from The Nottingham Longitudinal Study of Activity and Ageing (NLSAA) project.''
E. Turcan and K. McKeown, "Dreaddit: A Reddit Dataset for Stress Analysis in Social Media," in Proceedings of the Tenth International Workshop on Health Text Mining and Information Analysis (LOUHI 2019), Hong Kong, China, Nov. 2019, pp. 97–107. DOI: https://doi.org/10.18653/v1/D19-6213
A. R. Salehi and M. Khedmati, ''A cluster-based SMOTE both-sampling (CSBBoost) ensemble algorithm for classifying imbalanced data,'' Scientific Reports, vol. 14, no. 1, Mar. 2024, Art. no. 5152. DOI: https://doi.org/10.1038/s41598-024-55598-1
K. Liu, Y. Ye, S. Li, and H. Tang, ''Accurate de novo peptide sequencing using fully convolutional neural networks,'' Nature Communications, vol. 14, no. 1, Dec. 2023, Art. no. 7974. DOI: https://doi.org/10.1038/s41467-023-43010-x
Downloads
How to Cite
License
Copyright (c) 2025 Kamalamma Puttaraju Impana, Mullur Puttubuddhi Pushpalatha, Kotagi Basavarajappa Vikhyath, Rajanna Raksha
This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain the copyright and grant the journal the right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) after its publication in ETASR with an acknowledgement of its initial publication in this journal.
