Research on Anti-noise Speech Recognition System
碩士 === 國立成功大學 === 造船及船舶機械工程學系碩博士班 === 91 === The purpose of this research is to apply wavelet and statistics for spoken language process . To decide speech correct to start and stop point for recognize result，building up a set recognize system，speech waveform in time and recognize result to show in...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | zh-TW |
| Published: |
2003
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/26461859626164939962 |
| Summary: | 碩士 === 國立成功大學 === 造船及船舶機械工程學系碩博士班 === 91 === The purpose of this research is to apply wavelet and statistics for spoken language process . To decide speech correct to start and stop point for recognize result，building up a set recognize system，speech waveform in time and recognize result to show in the personal computer.
Use matlab program language in this research，collocation personal computer for operational interface and analogy to digital transform by sound card.
This paper present the recognize system，besides can performance in quiet environment，more enhance to every different S/N recognize for an objective，in other words is something of background noise in normal environment，however to be suitable for use by developed recognize system in purpose，even if to increase practicability by system.
The experimental results （from 0~9、password for operate ship）by wavelet 、statistics and energy. Speech recognition can be greatly improved and decrease time by using the new feature method under noise.
|
|---|