M. Shimizu, T. Fukushima, M. Dantsuji, and K. Shimizu (Japan)
database design, non-native speech database,multimedia-based speech database, parameters for facialmeasurement, non-invasive data acquisition
In recent times we have come to see a variety of non-native speech databases for use with speech recognition CALL (Computer-Assisted Language Learning) systems appropriate for non-native speakers. In this study, we describe a multimedia-based non-native speech database of English spoken by Japanese students currently under development in our laboratory. This database was designed not only for use with construction of phone models for English spoken by Japanese learners, but also for use with phonetic and linguistic analysis of English spoken by Japanese learners. In addition, we propose a non-invasive method for analyzing facial image data.
Important Links:
Go Back
