A study on efficient playing system with flexible time axis and speech database for researchers of speech

点击翻译按钮获取中文摘要

A method for reproducing a speech by the optimum reproduction speed has been studied based on hearing characteristics. This study consists of two subjects, one is quantitative analysis of relationship between speaking rate and recognition rate of speech, and another is implementation and experiments of reproducing system. The speaking rate variation speech database (SRV-DB) was also constructed and opened to the public for the researchers. These subjects have been completed according to the schedule.