The USF audiovisual phoneme database, v 1.0
Stefan A. Frisch, Sarah Hardin, Nikki Martin, Dee Adams Nikjeh, and Adrienne M. Stearns
University of South Florida

An audiovisual database of English speech sounds has been developed for 
teaching purposes. This database contains a wide variety of Standard 
English speech sounds produced in clear speech in isolated words. Phonemes 
are produced in word initial, word medial, and word final position, unless 
not allowed by English phonotactics. There is one example of each word 
spoken by a male talker of Standard American English from the Midwest. The 
database consists of individual word files that contain simultaneous audio 
recording, video of the lips and jaw, flexible endoscopic image of the 
pharynx and larynx, and ultrasound video of the tongue in the mid-saggital 
plane. The files in the database are uncompressed video avi format and are 
suitable for examination in the Wavesurfer freeware program in audio or 
video modes (Sjolander and Beskow, KTH Stockholm). This database is 
intended as a multimedia reference for students in phonetics or speech 
science. Its coverage is relatively complete, though there are some gaps 
due to technical difficulties with the recording procedure. A 
demonstration of selected recordings from the database and suggestions for 
their use will be presented. Plans for further development of the database 
will also be discussed.