Title:
VOICE RECOGNITION METHOD, APPARATUS, COMPUTER DEVICE, AND STORAGE MEDIUM
Document Type and Number:
WIPO Patent Application WO/2018/227781
Kind Code:
A1
Abstract:
The present application provides a method for voice recognition, said method comprising: obtaining to-be-recognized voice data; extracting a filter-bank feature and MFCC feature from voice data; taking the MFCC feature to be input data of a GMM-HMM model, and obtaining a first likelihood probability matrix; taking the filter-bank feature to be an input feature of a two-dimensional LSTM model, and obtaining a posterior probability matrix; taking the posterior probability matrix and first likelihood probability matrix to be input data of an HMM model, obtaining a second likelihood probability matrix, and according to the second likelihood probability matrix, obtaining a corresponding target word sequence from a phoneme decoding network.
Inventors:
LIANG HAO (CN)
WANG JIANZONG (CN)
CHENG NING (CN)
XIAO JING (CN)
WANG JIANZONG (CN)
CHENG NING (CN)
XIAO JING (CN)
Application Number:
PCT/CN2017/100049
Publication Date:
December 20, 2018
Filing Date:
August 31, 2017
Export Citation:
Assignee:
PING AN TECH SHENZHEN CO LTD (CN)
International Classes:
G10L15/02; G10L15/14
Foreign References:
| CN105976812A | 2016-09-28 | |||
| CN106557809A | 2017-04-05 | |||
| CN105206258A | 2015-12-30 |
Other References:
HSU, WEI-NING ET AL.: "A prioritized grid long short-term memory RNN for speech recognition", IEEE PROC. 2016 SPOKEN LANGUAGE TECHNOLOGY WORKSHOP (SLT), 13 December 2016 (2016-12-13) - 16 December 2016 (2016-12-16), San Diego, California, pages 467 - 473, XP033061780, DOI: 10.1109/SLT.2016.7846305
LI,JINYU ET AL.: "Exploring multidimensional LSTMS for large vocabulary ASR", IEEE WORKSHOP ON AUTOMATIC SPEECH RECOGNITION AND UNDERSTANDING (ASRU), 2016, pages 4940 - 4944, XP032901543, DOI: 10.1109/ICASSP.2016.7472617
LI,JINYU ET AL.: "Exploring multidimensional LSTMS for large vocabulary ASR", IEEE WORKSHOP ON AUTOMATIC SPEECH RECOGNITION AND UNDERSTANDING (ASRU), 2016, pages 4940 - 4944, XP032901543, DOI: 10.1109/ICASSP.2016.7472617
Attorney, Agent or Firm:
ADVANCE CHINA IP LAW OFFICE (CN)
Download PDF: