We present computer tools to help speech-language pathologists (SLP) in counting disfluencies, for both real-time and transcriptbased counts. The latter tend to be more precise and show which words are involved in each disfluency. Our approach allows realtime counts to be used as the basis for transcript-based counts. We employ automatic speech recognition to generate a word transcript (for read-speech samples), and then automatically merge the disfluency annotations with the word transcript, and have the SLP review parts of the audio file where a disfluency annotation was placed. This approach results in improved disfluency annotations, and can be done in 3 times real-time.
Bibliographic reference. Heeman, Peter A. / McMillin, Andy / Yaruss, J. Scott (2011): "Computer-assisted disfluency counts for stuttered speech", In INTERSPEECH-2011, 3013-3016.