Disfluency is the first phenomenon one has to address when processing spontaneous speech. Efficient systems combining transcription-based and signal-based cues have been created for English. These systems generally use supervised machine learning models, trained over large annotated datasets combining signal and transcription. As for other languages, including French, the situation is complicat...

Conversational speech from over 300 speakers from 17 to 68 years of age was analyzed for age-related changes in the timing and content of spoken language production. Overall, several relationships between the lexical content, timing, and fluency of speech emerged, such that more novel and lower frequency words were associated with slower speech and higher levels of disfluencies. Speaker age was...

The paper considers self-addressed queries – queries speakers address to themselves in the aftermath of a filled pause. We study their distribution in the BNC and show that such queries show signs of sensitivity to the syntactic/semantic type of the sub-utterance they follow. We offer a formal model that explains the coherence of such queries.

Most automatic speech recognition work has concentrated on read speech, whose acoustic aspects differ significantly from speech found in actual dialogues. A primary difference between read speech and spontaneous speech concerns a high rate of disfluencies (e.g., filled pauses, repetitions, repairs, false starts). Filled pauses (e.g., “uh,” “um”), unlike silences, resemble phones as part of word...

In an earlier paper, we have described final part-word repetitions in the conversational speech of two school-age boys of normal intelligence with no known neurological lesions. In this paper we explore in more detail the phonetic and linguistic characteristics of the speech of the boys. The repeated word fragments were more likely to be preceded by a pause than followed by one. The word immedi...

Automatically recognising children’s speech is a very difficult task. This difficulty can be attributed to the high variability in children’s speech, both within and across speakers. The variability is due to developmental changes in children’s anatomy, speech production skills et cetera, and manifests itself, for example, in fundamental and formant frequencies, the frequency of disfluencies, a...

This work explores prosodic cues of disfluencies in a corpus of university lectures. Results show three significant (p < 0.001) trends: pitch and energy slopes are significantly different between the disfluency and the onset of fluency; those features are also relevant to disfluency type differentiation; and they do not seem to be a speakereffect. The best combination of linguistic features one...

In this paper, we present a Conditional Random Field based approach for automatic detection of edit disfluencies in a conversational telephone corpus in French. We define disfluency patterns using both linguistic and acoustic features to perform disfluency detection. Two related tasks are considered : the first task aims at detecting the disfluent speech portion proper or reparandum, i.e. the p...

Automatic detection of disfluencies in spoken language is important for making speech recognition output more readable, and for aiding downstream language processing modules. We compare a generative hidden Markov model (HMM)-based approach and two conditional models — a maximum entropy (Maxent) model and a conditional random field (CRF) — for detecting disfluencies in speech. The conditional mo...

This paper presents an ASS (Automatic Speech Segmentation) Technique to segment spontaneous speech into syllable like units. In the development of a syllable-centric ASS system, segmentation of the acoustic signal into syllabic units is an important stage. In this paper we focus on the identifying minimum unit of speech to be considered while training any speech recognition system. There are sy...