Patent
Method for speech recognition system improving discrimination by using sampling velocity conversion
Zhenhua Huang,Limin Hou +1 more
- 10 Dec 2008
10
TL;DR: In this paper, the sampling rate is normalized before the speech is recognized, so that the sampling rates of the testing speech and the training speech are consistent to reduce the mistaken recognition rate as a result of inconsistent sampling rates.
read more
Abstract: The present invention relates to a method for improving the recognition rate of a speech recognition system through the sampling rate conversion. The method is that: the sampling rate is normalized before the speech is recognized, so that the sampling rates of the testing speech and the training speech are consistent to reduce the mistaken recognition rate as a result of inconsistent sampling rates. Moreover, the method proposes a novel sampling rate conversion structure of a fraction sampling rate conversion. Compared with the prior art, the method can completely reserve the original frequency spectrum, and prevent partly loss of the high-frequency part of the frequency spectrum when the sampling rate is converted. Therefore, the method can effectively improve the mistaken recognition rate as a result of the inconsistent sampling rates in the isolated word recognition system of large vocabulary.
read more
Chat with Paper
AI Agents for this Paper
Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps
Citations
Patent
Subtitle synchronization device and subtitle synchronization method based on speech recognition
TL;DR: In this article, a speech recognition module is used for extracting speech in foreground voice from an audio stream, and sampling and recognizing the extracted speech to generate corresponding text information; the dynamic sampling adjustment module was used for evaluating the degree of semantic recognition of the generated text information.
19
Patent
Automatic testing method and automatic testing device
Ning Jing,Yuan Jianfa +1 more
- 12 Feb 2014
TL;DR: In this article, an automatic testing method and an automatic device are used for performing automatic testing on a voice recognition system using pre-stored text files, which is suitable for large number of voice recognition tests and has the advantages of saving manpower resource and being high in testing efficiency.
17
Patent
Voice recognition method and device
Duan Hong,Tang Liliang,Xie Yan,Peng Shouye +3 more
- 16 Sep 2015
TL;DR: In this article, a method consisting of firstly, acquiring the voice information input by a speaker and acquiring the information of the speaker, secondly, judging the existence/absence of a personal acoustic model corresponding to the speaker according to the information provided by the speaker and thirdly, upon judging the presence of the personal model, acquiring personal acoustic models and conducting the voice recognition on the voice input on the input according to a personal audio model, fourthly, generating the corpus information of a speaker based on the audio input, and fifthly, creating a personalized acoustic model for the
11
Patent
User-defined instruction recognition speech photographing system
Wang Dandan,Zang Xian +1 more
- 07 Sep 2016
TL;DR: In this article, a user-defined instruction recognition speech photographing system consisting of a speech instruction collecting module, an audio signal preprocessing module, audio signal feature extraction module, a speech definition training module and a speech recognition control module is described.
8
Patent
Terminal and server of speaker-adaptation speech-recognition system and method for operating the system
Dong Hyun Kim
- 11 May 2015
TL;DR: In this paper, a speaker-adaptation speech-recognition system is described, which includes a speech recorder which transmits speech data of a speaker to a speech recognition server, and a statistical variable accumulator which receives a statistical information about speech of the speaker from the speech-recording server.
6