https://scholars.lib.ntu.edu.tw/handle/123456789/630389
標題: | DON'T SPEAK TOO FAST: THE IMPACT OF DATA BIAS ON SELF-SUPERVISED SPEECH MODELS | 作者: | Meng, Yen Chou, Yi Hui Liu, Andy T. HUNG-YI LEE |
關鍵字: | Data Bias | Self-supervised Speech Models | SUPERB Benchmark | 公開日期: | 1-一月-2022 | 卷: | 2022-May | 來源出版物: | ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings | 摘要: | Self-supervised Speech Models (S3Ms) have been proven successful in many speech downstream tasks, like ASR. However, how pretraining data affects S3Ms' downstream behavior remains an unexplored issue. In this paper, we study how pre-training data affects S3Ms by pre-training models on biased datasets targeting different factors of speech, including gender, content, and prosody, and evaluate these pre-trained S3Ms on selected downstream tasks in SUPERB Benchmark. Our experiments show that S3Ms have tolerance toward gender bias. Moreover, we find that the content of speech has little impact on the performance of S3Ms across downstream tasks, but S3Ms do show a preference toward a slower speech rate. |
URI: | https://scholars.lib.ntu.edu.tw/handle/123456789/630389 | ISBN: | 9781665405409 | ISSN: | 15206149 | DOI: | 10.1109/ICASSP43922.2022.9747897 |
顯示於: | 電機工程學系 |
在 IR 系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。