A prosody-based approach to end-of-utterance detection that does not require speech recognition.

Luciana Ferrer, Elizabeth Shriberg, Andreas Stolcke

Browse the full ICASSP paper archive.