Review of the anti-noise method in the speech recognition technology

With the development of science and technology, the requirement in the aspect of dealing with the relationship between human and machine is higher and higher. Speech recognition is one subject based on the pattern recognition and is closely related to computer science, psychology, linguistics, and s...

Full description

Saved in:
Bibliographic Details
Published in2017 12th IEEE Conference on Industrial Electronics and Applications (ICIEA) pp. 1391 - 1394
Main Authors Liu, Jing, Xiang, Xinguang
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.06.2017
Subjects
Additive noise
anti-noise method
feature compensation technology
Hidden Markov models
Model compensation technology
Noise measurement
Speech
Speech enhancement
speech enhancement algorithm
Speech recognition
speech recognition technology
Online AccessGet full text
ISSN2158-2297
DOI10.1109/ICIEA.2017.8283056

Cover

More Information
Summary:With the development of science and technology, the requirement in the aspect of dealing with the relationship between human and machine is higher and higher. Speech recognition is one subject based on the pattern recognition and is closely related to computer science, psychology, linguistics, and signal processing, etc. The key step of speech recognition is the pretreatment process of speech signal. This paper introduces three common anti-noise speech processing technology, i.e., speech enhancement algorithm, feature compensation technology, and model compensation technology. Finally, some existing problems are enumerated in the speech recognition technology.
ISSN:2158-2297
DOI:10.1109/ICIEA.2017.8283056