https://scholars.lib.ntu.edu.tw/handle/123456789/580902
標題: | DARTS-ASR: Differentiable architecture search for multilingual speech recognition and adaptation | 作者: | Chen Y.-C Hsu J.-Y Lee C.-K Lee H.-Y. HUNG-YI LEE |
關鍵字: | Architecture; Speech communication; Topology; Character error rates; Fixed topologies; Gradient based; Hyperparameters; Model architecture; Multilingual speech recognition; Relative reduction; Speech recognition | 公開日期: | 2020 | 卷: | 2020-October | 起(迄)頁: | 1803-1807 | 來源出版物: | Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH | 摘要: | In previous works, only parameter weights of ASR models are optimized under fixed-topology architecture. However, the design of successful model architecture has always relied on human experience and intuition. Besides, many hyperparameters related to model architecture need to be manually tuned. Therefore in this paper, we propose an ASR approach with efficient gradient-based architecture search, DARTS-ASR. In order to examine the generalizability of DARTS-ASR, we apply our approach not only on many languages to perform monolingual ASR, but also on a multilingual ASR setting. Following previous works, we conducted experiments on a multilingual dataset, IARPA BABEL. The experiment results show that our approach outperformed the baseline fixed-topology architecture by 10.2% and 10.0% relative reduction on character error rates under monolingual and multilingual ASR settings respectively. Furthermore, we perform some analysis on the searched architectures by DARTS-ASR. Copyright ? 2020 ISCA |
URI: | https://www.scopus.com/inward/record.uri?eid=2-s2.0-85097970691&doi=10.21437%2fInterspeech.2020-1315&partnerID=40&md5=bbf43cc58caf01401ad2eb9ab323f984 https://scholars.lib.ntu.edu.tw/handle/123456789/580902 |
ISSN: | 2308457X | DOI: | 10.21437/Interspeech.2020-1315 |
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。