Articles

Distant Speech Recognition in a Smart Home : Comparison of Several Multisource ASRs in Realistic Conditions While the smart home domain has become a major field of application of ICT to improve support and wellness of people in loss of autonomy, speech technology in smart home has, comparatively to other …...

_________________________________________________________________________________________________________ Comparison of journalistic and spontaneous speech: analysis of sequences between pauses. In this study we compare the ESTER corpus of journalistic speech (Galliano et al., 2005) and the NCCF corpus of spontaneous speech (Torreira et al, 2010) in terms of duration, f0 and spectral reduction in productions automatically detected …...

Pronunciation generation for proper names using Conditional Random Fields We propose an approach to grapheme-to-phoneme conversion for proper names based on a probabilistic method: Conditional Random Fields (CRFs). CRFs give a long term prediction, assume a relafex state independence condition and allow a tag integration. In previous work, grapheme-to-phoneme conversion …...

__________________________________________________________________________________________________________ Utilization of the piezoelectric accelerometer for the study of nasality in French as a Foreign Language The aim of this paper is to compare the production of French nasal vowels pronounced by 5 French native speakers and 5 Bosnian learners of French. In Bosnian, phonological nasal vowels do not …...

Developments of Swahili resources for an automatic speech recognition system This article describes our efforts to provide ASR resources for Swahili, a Bantu language spoken in a wide area of East Africa. We start with an...

_________________________________________________________________________________________________________ Vowel lengthening in Belgium French: an experimental and perceptual approach This paper investigates the lengthening of certain vowels in Belgian French and its influence on the perception of a Belgian accent. Based on recordings made in Belgium, two perceptual experiments were conducted, involving Belgian and French listeners: the first …...

_________________________________________________________________________________________________________ Vocalic system’s typology revisited from the functional load viewpoint Most studies in phonological typology implicitly assume that all segments in an inventory are equally important. We suggest here that the notion of functional load enriches this usual representation of phonological systems. Focusing on the vowel systems of 12 languages, …...

_________________________________________________________________________________________________________ Corsican questions: is there a prosodic transfer from Corsican to French? This study investigates whether a prosodic transfer can be highlighted from Corsican (an Italo-Romance language) to French spoken in Corsica, where French is now the dominant language. A corpus of transparent sentences such as la touriste trouve la …...

__________________________________________________________________________________________________________ PROSOTRAN : an tool that provides a symbolic representation of the prosodic events in non-standard data Most of the existing prosodic transcription systems display some limitations: (i) they cannot account for the various prosodic phenomena (accentuation, phrasing and tonal patterns), as they focus on the variation of intonational patterns; …...

Source separation by cepstral smoothing of binary masks In this paper, we propose a separation system of speech signals from two convolutive mixtures. The suggested system is based on the combination of blind source separation technique with a time-frequency masking procedure, followed by a smoothing cepstral. Indeed, after separation of …...

_______________________________________________________________________________________________ Read the tones on the lips : visual perception(s) of lexical tones in Mandarin Chinese The aim of the present study is to verify whether the visual cues located on the neck, can contribute in Mandarin tones visual perception. However, in an unexpected way, this study shows that tones …...

_________________________________________________________________________________________________________ F2/F3 of voiced plosives and F1/F2 of vowel in VCV Sequences in Children with a Cleft Palate-An acoustic Study The aim of this acoustic study is to examine place of articulation during the production of voiced plosives by children with a cleft palate. The data are compared with those …...

Segmental evaluation of HTS HTS is a parametric speech synthesis system based on the use of Hidden Markov Models (HMM). HTS is now widely used but very few studies have been conducted to analyze the influence of parameters on the quality of the synthetic speech. The aim of this paper …...

Automatic measurement of prosodic accommodation in conversational interaction It has been observed in many studies that speakers, over the course of a conversation, adapt their verbal and non-verbal behaviour (lexicon, syntax, prosody, postures, gesture) to their interlocutor. This accommodation facilitates, on the one hand, the exchange of information, mutual understanding …...

COSMO, a Bayesian model of speech communication, applied to syllable perception This works uses the computational modeling framework to compare motor, auditory, and sensorimotor theories of speech communication. More precisely, we define a Bayesian model of a communicating agent which we simulate to carry out syllable perception tasks according to …...

Prosodic variations in unit-based speech synthesis: the example of interrogative sentences This paper proposes an automatic method to increase the number of possible prosodic variations in non-uniform unit-based speech synthesis. More specifically, we are interested in the production of interrogative sentences through the eLite text-to-speech synthesis system, which relies on …...

__________________________________________________________________________________________________________ Schwa elision in children-parental interactions: A corpus study The present article investigates schwa elision in the acquisition of French L1. Using four dense corpora of natural children-parental interactions, we examine the variation between the occurrence and non-occurrence of the schwa in both the children’s and adults’ speech. In particular, …...

_________________________________________________________________________________________________________ French Cued Speech perception and expertise effect in hearing people Since French Cued Speech (CS) is multi-signal (lip movements and CS gestures), we conducted an eye tracking study to examine whether this perception involves integrative treatment and how expertise affects it in normally-hearing participants. Our paradigm consisted in a …...

_________________________________________________________________________________________________________ Temporal dynamics of binding in audiovisual speech fusion The McGurk effect demonstrates the phenomenon of audiovisual fusion: a sound "ba" mounted on a video "ga" is often perceived as "da". In a previous work we showed that audiovisual fusion might be modulated by a precedent binding process (Nahorna et …...

Error region detection and characterization in transcriptions of multimedia documents : application to person name search In this article, we propose to detect and characterize error regions in automatic transcriptions of multimedia documents. The simultaneous detection and characterization could be seen as a sequence labeling task where we compare sequential …...