Depression recognition using voice-based pre-training model
Abstract The early screening of depression is highly beneficial for patients to obtain better diagnosis and treatment. While the effectiveness of utilizing voice data for depression detection has been demonstrated, the issue of insufficient dataset size remains unresolved. Therefore, we propose an a...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2024-06-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-024-63556-0 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841544596717305856 |
|---|---|
| author | Xiangsheng Huang Fang Wang Yuan Gao Yilong Liao Wenjing Zhang Li Zhang Zhenrong Xu |
| author_facet | Xiangsheng Huang Fang Wang Yuan Gao Yilong Liao Wenjing Zhang Li Zhang Zhenrong Xu |
| author_sort | Xiangsheng Huang |
| collection | DOAJ |
| description | Abstract The early screening of depression is highly beneficial for patients to obtain better diagnosis and treatment. While the effectiveness of utilizing voice data for depression detection has been demonstrated, the issue of insufficient dataset size remains unresolved. Therefore, we propose an artificial intelligence method to effectively identify depression. The wav2vec 2.0 voice-based pre-training model was used as a feature extractor to automatically extract high-quality voice features from raw audio. Additionally, a small fine-tuning network was used as a classification model to output depression classification results. Subsequently, the proposed model was fine-tuned on the DAIC-WOZ dataset and achieved excellent classification results. Notably, the model demonstrated outstanding performance in binary classification, attaining an accuracy of 0.9649 and an RMSE of 0.1875 on the test set. Similarly, impressive results were obtained in multi-classification, with an accuracy of 0.9481 and an RMSE of 0.3810. The wav2vec 2.0 model was first used for depression recognition and showed strong generalization ability. The method is simple, practical, and applicable, which can assist doctors in the early screening of depression. |
| format | Article |
| id | doaj-art-81d495a9cda74eafa45e74f4b8e90b5c |
| institution | Kabale University |
| issn | 2045-2322 |
| language | English |
| publishDate | 2024-06-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-81d495a9cda74eafa45e74f4b8e90b5c2025-01-12T12:25:10ZengNature PortfolioScientific Reports2045-23222024-06-0114111310.1038/s41598-024-63556-0Depression recognition using voice-based pre-training modelXiangsheng Huang0Fang Wang1Yuan Gao2Yilong Liao3Wenjing Zhang4Li Zhang5Zhenrong Xu6School of Biomedical Engineering, South-Central Minzu UniversitySchool of Biomedical Engineering, South-Central Minzu UniversitySchool of Biomedical Engineering, South-Central Minzu UniversitySchool of Biomedical Engineering, South-Central Minzu UniversitySchool of Biomedical Engineering, South-Central Minzu UniversitySchool of Biomedical Engineering, South-Central Minzu UniversitySchool of Biomedical Engineering, South-Central Minzu UniversityAbstract The early screening of depression is highly beneficial for patients to obtain better diagnosis and treatment. While the effectiveness of utilizing voice data for depression detection has been demonstrated, the issue of insufficient dataset size remains unresolved. Therefore, we propose an artificial intelligence method to effectively identify depression. The wav2vec 2.0 voice-based pre-training model was used as a feature extractor to automatically extract high-quality voice features from raw audio. Additionally, a small fine-tuning network was used as a classification model to output depression classification results. Subsequently, the proposed model was fine-tuned on the DAIC-WOZ dataset and achieved excellent classification results. Notably, the model demonstrated outstanding performance in binary classification, attaining an accuracy of 0.9649 and an RMSE of 0.1875 on the test set. Similarly, impressive results were obtained in multi-classification, with an accuracy of 0.9481 and an RMSE of 0.3810. The wav2vec 2.0 model was first used for depression recognition and showed strong generalization ability. The method is simple, practical, and applicable, which can assist doctors in the early screening of depression.https://doi.org/10.1038/s41598-024-63556-0DepressionPre-training modelVoice featuresWav2vec 2.0DAIC-WOZ |
| spellingShingle | Xiangsheng Huang Fang Wang Yuan Gao Yilong Liao Wenjing Zhang Li Zhang Zhenrong Xu Depression recognition using voice-based pre-training model Scientific Reports Depression Pre-training model Voice features Wav2vec 2.0 DAIC-WOZ |
| title | Depression recognition using voice-based pre-training model |
| title_full | Depression recognition using voice-based pre-training model |
| title_fullStr | Depression recognition using voice-based pre-training model |
| title_full_unstemmed | Depression recognition using voice-based pre-training model |
| title_short | Depression recognition using voice-based pre-training model |
| title_sort | depression recognition using voice based pre training model |
| topic | Depression Pre-training model Voice features Wav2vec 2.0 DAIC-WOZ |
| url | https://doi.org/10.1038/s41598-024-63556-0 |
| work_keys_str_mv | AT xiangshenghuang depressionrecognitionusingvoicebasedpretrainingmodel AT fangwang depressionrecognitionusingvoicebasedpretrainingmodel AT yuangao depressionrecognitionusingvoicebasedpretrainingmodel AT yilongliao depressionrecognitionusingvoicebasedpretrainingmodel AT wenjingzhang depressionrecognitionusingvoicebasedpretrainingmodel AT lizhang depressionrecognitionusingvoicebasedpretrainingmodel AT zhenrongxu depressionrecognitionusingvoicebasedpretrainingmodel |