Low-latency convolutional recurrent neural network for keyword spotting

Hu Du, Ruohan Li, Donggyun Kim, Kaoru Hirota, Yaping Dai

科研成果: 书/报告/会议事项章节会议稿件同行评审

4 引用 (Scopus)

摘要

A Low-latency Convolutional Recurrent Neural Network (L-CRNN) is proposed to reduce the complexity of a Keyword Spotting (KWS) system with high accuracy. The L-CRNN reduces a number of parameters between RNN layer and Full-Connected (FC) layer, which saves at least 1/2 memory for on-hands device compared with Convolutional Recurrent Neural Network (CRNN) depending on the number of FC units. Furthermore, it learns valid deep audio features to classify the keywords and garbage words with high accuracy. Results of experiments on the Google's Speech Commands Datasets show that the L-CRNN achieves 96.17% accuracy with less than 1/4 number of parameters and fewer float operations compared with Convolutional Neural Network (CNN) and CRNN.

源语言英语
主期刊名Proceedings - 2018 Joint 10th International Conference on Soft Computing and Intelligent Systems and 19th International Symposium on Advanced Intelligent Systems, SCIS-ISIS 2018
出版商Institute of Electrical and Electronics Engineers Inc.
802-807
页数6
ISBN(电子版)9781538626337
DOI
出版状态已出版 - 2 7月 2018
活动Joint 10th International Conference on Soft Computing and Intelligent Systems and 19th International Symposium on Advanced Intelligent Systems, SCIS-ISIS 2018 - Toyama, 日本
期限: 5 12月 20188 12月 2018

出版系列

姓名Proceedings - 2018 Joint 10th International Conference on Soft Computing and Intelligent Systems and 19th International Symposium on Advanced Intelligent Systems, SCIS-ISIS 2018

会议

会议Joint 10th International Conference on Soft Computing and Intelligent Systems and 19th International Symposium on Advanced Intelligent Systems, SCIS-ISIS 2018
国家/地区日本
Toyama
时期5/12/188/12/18

指纹

探究 'Low-latency convolutional recurrent neural network for keyword spotting' 的科研主题。它们共同构成独一无二的指纹。

引用此