The Development of Robust Speech Recognition system based on iOS devices
碩士 === 元智大學 === 通訊工程學系 === 101 === The thesis presents the robust techniques for resource-constrained speech recognition (SR) on iOS mobile devices. Among the related issues and techniques we explore are: (1) A POSIX threads-based framework is designed for SR on iOS devices and thus a significant re...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | zh-TW |
| Online Access: | http://ndltd.ncl.edu.tw/handle/91978061275964550941 |
| Summary: | 碩士 === 元智大學 === 通訊工程學系 === 101 === The thesis presents the robust techniques for resource-constrained speech recognition (SR) on iOS mobile devices. Among the related issues and techniques we explore are: (1) A POSIX threads-based framework is designed for SR on iOS devices and thus a significant reducing of total response time is obtained. (2) A novel UED (Utterance-End Detection) is integrated with SR to eliminate redundant computation. (3) The robust acoustic models are incorporated in SR for achieving good performances in noisy environments. Finally, our experimental results indicate that the proposed methods are operative for SR on iOS devices.
|
|---|