Authors:
Piotr Żelasko
1
;
Tomasz Jadczyk
2
and
Bartosz Ziółko
2
Affiliations:
1
AGH University of Science and Technology, Poland
;
2
AGH University of Science and Technology and Techmo, Poland
Keyword(s):
ASR, Breath, Breath Detection, Filled Pause, Filled Pause Detection, Filler, Filler detection, HMM, IVR, Speech, Speech Recognition, Spontaneous Speech.
Related
Ontology
Subjects/Areas/Topics:
Biometrics and Pattern Recognition
;
Design and Implementation of Signal Processing Systems
;
Human-Machine Interface
;
Multimedia
;
Multimedia Signal Processing
;
Multimedia Systems and Applications
;
Telecommunications
Abstract:
The phenomena of filled pauses and breaths pose a challenge to Automatic Speech Recognition (ASR) systems dealing with spontaneous speech, including recognizer modules in Interactive Voice Reponse (IVR) systems. We suggest a method based on Hidden Markov Models (HMM), which is easily integrated into HMM-based ASR systems and allows detection of those disturbances without incorporating additional parameters. Our method involves training the models of disturbances and their insertion in the phrase Markov chain between word-final and word-initial phoneme models. Application of the method in our ASR shows improvement of recognition results in Polish telephonic speech corpus LUNA.