发明授权
- 专利标题: Full-sequence training of deep structures for speech recognition
- 专利标题(中): 用于语音识别的深层结构的全序训练
-
申请号: US12886568申请日: 2010-09-21
-
公开(公告)号: US09031844B2公开(公告)日: 2015-05-12
- 发明人: Dong Yu , Li Deng , Abdel-rahman Samir Abdel-rahman Mohamed
- 申请人: Dong Yu , Li Deng , Abdel-rahman Samir Abdel-rahman Mohamed
- 申请人地址: US WA Redmond
- 专利权人: Microsoft Technology Licensing, LLC
- 当前专利权人: Microsoft Technology Licensing, LLC
- 当前专利权人地址: US WA Redmond
- 代理商 Sandy Swain; Judy Yee; Micky Minhas
- 主分类号: G10L15/14
- IPC分类号: G10L15/14 ; G06K9/62 ; G06N3/04 ; G06N3/08
摘要:
A method includes an act of causing a processor to access a deep-structured model retained in a computer-readable medium, the deep-structured model includes a plurality of layers with respective weights assigned to the plurality of layers, transition probabilities between states, and language model scores. The method further includes the act of jointly substantially optimizing the weights, the transition probabilities, and the language model scores of the deep-structured model using the optimization criterion based on a sequence rather than a set of unrelated frames.
公开/授权文献
信息查询