| Summary: | 碩士 === 中原大學 === 資訊工程研究所 === 89 === This research develops a speech recognition technology for controlling the Personal Digital Assistant (PDA). The system constructed on Pocket PC platform with Windows CE as operating system. The system use speech recognition technology to assistant users executing the command on PDA.
After analogy to digital convertor, the digitized speech signals are obtained. Using energy detection and zero crossing technology, the available part of speech signals are extracted. The system utilize the coefficient of Mel-Frequency Cepstrum as the feature parameter, then use the method of Dynamic Time Warping to compare the speech input feature with the sample feature in the speech data file. The recognized instruction can be implemented by using program interface of Windows Application Programming Interface (API) and Pocket Outlook Object Model (POOM). The functions of the system include inquiry, making appointments and exploring Internet, etc. By using speech input command the operating of PDA is more convenient.
The performance of the speech control system is evaluated, the recognition rate under the normal operating environment is 86% and each speech input takes 1.085 second as recognition time.
|
|---|
