MF$^{2}$-Net: Exploring a Meta-Fuzzy Multimodal Fusion Network for Depression Recognition
- 13 May 2025
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Fuzzy Systems
- Vol. 33 (9) , 2924-2936
Abstract
Depression is a prevalent mental illness that significantly impacts the well-being of individuals and the development of society. The current diagnostic methods are largely subjective and time-consuming. Moreover, the existing machine learning-based depression recognition methods struggle to fully exploit the collaborative benefits between modalities, lack interpretability in their fusion processes, and perform inadequately in few-shot depression recognition tasks. To address these challenges, we propose a meta-fuzzy multimodal fusion network (MF$^{2}$-Net) for depression recognition. This innovative approach integrates physiological signals and behavioral data, employs multiple MLPs to learn the fuzzy measures of single base learners and complementary increments, then constructs all fuzzy measures, and finally achieves an interpretable decision-level fusion process through fuzzy integrals. Furthermore, we incorporate model-agnostic meta-learning for conducting few-shot domain-adaptive training, mitigating the issues related to high individual variability levels and the scarcity of multimodal depression data. Our method demonstrates exceptional classification performance in subject-independent experiments implemented on public datasets; offers a reliable solution for objectively, effectively, and conveniently recognizing depression; and has the potential to promote the clinical applications of rapid intelligent depression diagnosis.Keywords
Funding Information
- National Natural Science Foundation of China (62373057, 62227807, 62072219)
- National Key R&D Program of China (2019YFA0706200)
- Project Funded by China Postdoctoral Science Foundation (2021M700423, 2024T171131, 2024M764142)
- China University Research Innovation Fund (2024LC016)
- Scientific and Technological Innovation 2030 Project of MOST (2021ZD0201900, 2021ZD0200601)
This publication has 45 references indexed in Scilit:
- Few-shot classification via efficient meta-learning with hybrid optimizationEngineering Applications of Artificial Intelligence, 2023
- Few-shot and meta-learning methods for image understanding: a surveyInternational Journal of Multimedia Information Retrieval, 2023
- Task-aware meta-learning paradigm for universal structural damage segmentation using limited imagesEngineering Structures, 2023
- A multimodal fusion model with multi-level attention mechanism for depression detectionBiomedical Signal Processing and Control, 2022
- Content-based multiple evidence fusion on EEG and eye movements for mild depression recognitionComputer Methods and Programs in Biomedicine, 2022
- DepHNN: A novel hybrid neural network for electroencephalogram (EEG)-based screening of depressionBiomedical Signal Processing and Control, 2021
- Learning Fuzzy-Valued Fuzzy Measures for the Fuzzy-Valued Sugeno Fuzzy Integral2010
- Estimation of adult skeletal age‐at‐death using the Sugeno fuzzy integralAmerican Journal of Physical Anthropology, 2009
- The elements of statistical learning: data mining, inference and predictionThe Mathematical Intelligencer, 2005
- A Mathematical Theory of CommunicationBell System Technical Journal, 1948