Satoshi Nakamura Bücher

The book introduces a comprehensive statistical method for modeling the pronunciation of non-native speakers, highlighting the variations in their speech compared to native speakers. It addresses common phonetic deviations such as substitutions, deletions, and insertions. This innovative approach allows for automatic modeling of these pronunciation differences, providing valuable insights into second-language speech patterns.

A Speech Translation is a dream technology of all of the people living on the Globe. Recent advancement in transportations and communications has brought us borderless strong ties in economic and cultural relations worldwide. ATR (Advanced Telecommunications Research Institute International) has been pioneering Speech-to-Speech Translation technologies since 1986. This book is the result of intensive research of ATR scientists. In particular, it presents the yield of recent five-year projects on a corpus-based speech-to-speech translation system for travel conversation, funded by the Japanese government. The book shows the state-of-the-art technologies for speech translation and a wide variety of evaluation results, including field experiments conducted at the Kansai International Airport.