基本信息
周世玉  男  硕导  中国科学院自动化研究所
电子邮件: shiyu.zhou@ia.ac.cn
通信地址: 北京市海淀区中关村东路95号智能化大厦8层
邮政编码:

教育背景

2013-09--2019-01   中国科学院自动化研究所   博士
2007-09--2009-12   浙江大学   硕士
2002-09--2006-06   湖南师范大学   学士

专利与奖励

   
专利成果
( 1 ) 端到端的在线语音检测与识别方法、系统及设备, 发明专利, 2022, 第 1 作者, 专利号: CN112951213B

( 2 ) 低资源多语言的语音识别模型、语音识别方法, 发明专利, 2019, 第 1 作者, 专利号: CN110428818A

出版信息

   
发表论文
(1) CIF-BASED COLLABORATIVE DECODING FOR END-TO-END CONTEXTUAL SPEECH RECOGNITION, 2021 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP 2021), 2021, 第 3 作者
(2) Efficiently Fusing Pretrained Acoustic and Linguistic Encoders for Low-Resource Speech Recognition, IEEE SIGNAL PROCESSING LETTERS, 2021, 第 2 作者
(3) TWO-STAGE PRE-TRAINING FOR SEQUENCE TO SEQUENCE SPEECH RECOGNITION, IJCNN, 2021, 第 2 作者
(4) A Comparison of Modeling Units in Sequence-to-Sequence Speech Recognition with the Transformer on Mandarin Chinese, ICONIP, 2018, 第 4 作者
(5) Syllable-Based Sequence-to-Sequence Speech Recognition with the Transformer in Mandarin Chinese, Interspeech, 2018, 第 2 作者
(6) Extending Recurrent Neural Aligner for Streaming End-to-End Speech Recognition in Mandarin, 19TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION (INTERSPEECH 2018), VOLS 1-6, 2018, 第 2 作者
(7) 低资源语言的多语言语音识别建模方法研究, 2018, 第 1 作者
(8) A Comparison of Modeling Units in Sequence-to-Sequence Speech Recognition with the Transformer on Mandarin Chinese, NEURAL INFORMATION PROCESSING (ICONIP 2018), PT V, 2018, 通讯作者
(9) Word-level Permutation and Improved Lower Frame Rate for RNN-Based Acoustic Modeling, Iconip2017, 2017, 第 3 作者
(10) Multilingual Recurrent Neural Networks with Residual Learning for Low-Resource Speech Recognition, 18TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION (INTERSPEECH 2017), VOLS 1-6, 2017, 通讯作者
(11) Multilingual Recurrent Neural Networks with Residual Learning for Low-Resource Speech Recognition, Interspeech, 2017, 第 2 作者