International audienceSpontaneous speech may be characterized by a relatively high rate of disfluencies. Keeping disfluencies in transcriptions may help to better understand the relative progress in dialogue, however their presence may harm a smooth application of various automatic processings. In this contribution, we propose a method to automatic disfluency detection in conversational speech. To this aim, we implemented a CRF (Conditional Random Field) approach to tag our data into disfluent / non-disfluent regions. CRFs are known to perform quite well on this kind of segmentation / labeling tasks. Implemented features include both lexical, morpho-syntactic and acoustic-prosodic information. Contrastive experiments show that the addition ...
The aim of the PhD is to study in a detailed way the impact of disfluencies in spoken french (repeat...
We propose a novel algorithm to detect disfluency in speech by reformulating the problem as phrase-l...
In numerous domains, the usage of synthetic speech is conditioned upon the ability of speech synthes...
International audienceSpontaneous speech may be characterized by a relatively high rate of disfluenc...
International audienceSpontaneous speech may be characterized by a relatively high rate of disfluenc...
International audienceIn this paper, we present a Conditional Random Field based approach for automa...
International audienceIn this paper, we present a Conditional Random Field based approach for automa...
International audienceIn this paper, we present a Conditional Random Field based approach for automa...
In this paper we propose a multi-step system for the semiautomatic detection and annotation of disfl...
The disfluencies are a frequently occurring phenomenon in any spontaneous speech production; it cons...
The disfluencies are a frequently occurring phenomenon in any spontaneous speech production; it cons...
A key difference between spontaneous speech and controlled laboratory speech is the prevalence of di...
Extraire de l'information de données langagières est un sujet de plus en plus d'actualité compte ten...
National audienceThis paper presents an exploratory work on the automatic insertion of disfluencies ...
National audienceThis paper presents an exploratory work on the automatic insertion of disfluencies ...
The aim of the PhD is to study in a detailed way the impact of disfluencies in spoken french (repeat...
We propose a novel algorithm to detect disfluency in speech by reformulating the problem as phrase-l...
In numerous domains, the usage of synthetic speech is conditioned upon the ability of speech synthes...
International audienceSpontaneous speech may be characterized by a relatively high rate of disfluenc...
International audienceSpontaneous speech may be characterized by a relatively high rate of disfluenc...
International audienceIn this paper, we present a Conditional Random Field based approach for automa...
International audienceIn this paper, we present a Conditional Random Field based approach for automa...
International audienceIn this paper, we present a Conditional Random Field based approach for automa...
In this paper we propose a multi-step system for the semiautomatic detection and annotation of disfl...
The disfluencies are a frequently occurring phenomenon in any spontaneous speech production; it cons...
The disfluencies are a frequently occurring phenomenon in any spontaneous speech production; it cons...
A key difference between spontaneous speech and controlled laboratory speech is the prevalence of di...
Extraire de l'information de données langagières est un sujet de plus en plus d'actualité compte ten...
National audienceThis paper presents an exploratory work on the automatic insertion of disfluencies ...
National audienceThis paper presents an exploratory work on the automatic insertion of disfluencies ...
The aim of the PhD is to study in a detailed way the impact of disfluencies in spoken french (repeat...
We propose a novel algorithm to detect disfluency in speech by reformulating the problem as phrase-l...
In numerous domains, the usage of synthetic speech is conditioned upon the ability of speech synthes...