Sixth International Conference on Spoken Language Processing
(ICSLP 2000)

Beijing, China
October 16-20, 2000

Detection of Speech Landmarks Using Temporal Cues

Ariel Salomon, Carol Espy-Wilson

Electrical and Computer Engineering Department, Boston University, Boston, MA, USA

In order to improve the performance of speech recognizers, particularly in degraded environments, it may be beneficial to integrate use of temporal information. As literature has shown that human listeners are able to use temporal cues in speech recognition tasks, this study examines algorithms for extraction of temporal cues in a speech signal. The task under analysis is the location of landmarks to direct further analysis in a knowledge-based speech recognition system. In this study, a set of robust landmark types were located with an accuracy of 82.5%-100.0%.

Full Paper

Bibliographic reference.  Salomon, Ariel / Espy-Wilson, Carol (2000): "Detection of speech landmarks using temporal cues", In ICSLP-2000, vol.3, 762-765.