| Author: | Zhang, Tao |
| Title: | Spoken language processing and natural language processing for depression detection |
| Advisors: | Mak, Man Wai (EEE) |
| Degree: | M.Sc. |
| Year: | 2024 |
| Department: | Department of Electrical and Electronic Engineering |
| Pages: | v, 38 pages : color illustrations |
| Language: | English |
| Abstract: | Depression detection remains challenging due to the lack of standardized diagnostic methods, with clinical assessments by healthcare professionals being the primary approach. This method is inherently subjective and may not be reproducible. Current research in intelligent depression detection often focuses on isolated factors, which can result in a limited understanding of the condition. Our study enhanced a previous fusion model by incorporating pre-trained models to extract features from audio and text data. Specifically, we utilize the EATD-corpus and employ two pre-trained models: Wav2vec2.0 for audio feature extraction and RoBERTa for text feature extraction. Both models are fine-tuned using Chinese datasets to optimize their performance on the EATD Corpus. Additionally, we employ a deep neural network (DNN) model to combine and classify text and audio data into depressed and non-depressed data simultaneously. The final results of our model demonstrate a significant improvement in the extraction of relevant features from both audio and text data for depression detection. |
| Rights: | All rights reserved |
| Access: | restricted access |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| 8719.pdf | For All Users (off-campus access for PolyU Staff & Students only) | 1.91 MB | Adobe PDF | View/Open |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/14052