Speech-recorded event notification system

• Main Authors: Yu-Huan Liou, 劉宇桓
• Other Authors: Chung-Jen Kuo
• Format: Others
• Language: zh-TW
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/04374062166017803223

碩士 === 崑山科技大學 === 數位生活科技研究所 === 98 === Speech is one of a very convenient user interface. The study design of a Speech-recorded event notification system. As the future of embedded systems application-friendly User Interface.The system uses the time tag as the basic speech recognition result output unit when. For the user input to do the voice message content knowledge. We use the Hidden Markov Model Speech Recognition-based framework. Speech by the basic foundation of the country as a model sub-syllable units, using TCC300 corpus training. Identification of the consideration to the import of statements but time tags in addition, there are notes for the contents, therefore, the way we use the modified identification of the reference table, and Pei Ge phase search algorithm for identification. In the revised identification of the reference table, in addition to want to identify the time label, the follow-up will not in time for the rest of the syllable label use to be included, with a stage of the search algorithm, can be identified by the time to record labels. Allows users to tell the time and events Notepad words, after time for tag identification, the system will time-tag Jiang event Chucun good, when the system time is up Shi Yongzhe the preset time, the system will be a reminder for the event.