Résumé
Ce texte introductif présente le contexte dans lequel s’inscrit la recherche sur la phraséologie des interactions orales. Les dix articles qui composent le volume sont également synthétisés. ...Ces derniers regroupent des réflexions méthodologiques et des études de cas.
This article provides an overview of methodological and technical issues that arise in the collection, indexing and use of spoken learner corpora, i. e. corpora containing spoken utterances of ...learners of a target language. After an introductory discussion of the most important special features of this type of corpus that distinguish it from written language learner corpora and spoken corpora with L1 speakers, we will go into more detail on questions of corpus design. The main part of the paper is then an overview of the methodological and technical procedures of the individual steps of collecting, indexing, providing and using spoken learner corpora. The main aim of this overview is to highlight practices that can be considered best practices according to the current state of research. Finally, we outline the challenges that still exist for this type of corpus.
•Problem: end-to-end speech translation requires large corpora to train neural models.•Contribution: MuST-C is a large multilingual corpus built from English TED Talks.•Corpus content: English ...speech, aligned transcription/translations in 14 languages.•Other key features: high topic and speaker variety, large size, free distribution.•Discussion: empirical/manual quality evaluation, baseline results on all languages.
End-to-end spoken language translation (SLT) has recently gained popularity thanks to the advancement of sequence to sequence learning in its two parent tasks: automatic speech recognition (ASR) and machine translation (MT). However, research in the field has to confront with the scarcity of publicly available corpora to train data-hungry neural networks. Indeed, while traditional cascade solutions can build on sizable ASR and MT training data for a variety of languages, the available SLT corpora suitable for end-to-end training are few, typically small and of limited language coverage. We contribute to fill this gap by presenting MuST-C, a large and freely available Multilingual Speech Translation Corpus built from English TED Talks. Its unique features include: i) language coverage and diversity (from English into 14 languages from different families), ii) size (at least 237 hours of transcribed recordings per language, 430 on average), iii) variety of topics and speakers, and iv) data quality. Besides describing the corpus creation methodology and discussing the outcomes of empirical and manual quality evaluations, we present baseline results computed with strong systems on each language direction covered by MuST-C.
Éditorial – L’anglais oral Terrier, Linda
Recherche et pratiques pédagogiques en langues de spécialité,
2021, Letnik:
40, Številka:
1
Journal Article
Odprti dostop
C’est avec beaucoup d’honneur et un grand plaisir que Recherche et pratiques pédagogiques en langues de spécialité - Les cahiers de l’Apliut accueille pour la première fois dans ses pages un numéro ...coordonné par l’Aloes, l’association des anglicistes pour les études de langue orale dans l’enseignement supérieur, secondaire et primaire. L’idée de cette collaboration a émergé en 2017, lors des journées d’études de l’Aloes organisées par Susan Moore Mauroux à l’université de Limoges sur le thème...
•The general language network of the early deaf brain revealed by meta-analysis•Broca's area and the left pMTG robustly participate in supramodal language processing•More activation in the left ...calcarine gyrus and the right caudate in deaf individuals•Cross-modal plasticity and age of sign language acquisition shape language network•An overall left-lateralized pattern of language with regional lateralization shifts
This meta-analysis summarizes evidence from 44 neuroimaging experiments and characterizes the general linguistic network in early deaf individuals. Meta-analytic comparisons with hearing individuals found that a specific set of regions (in particular the left inferior frontal gyrus and posterior middle temporal gyrus) participates in supramodal language processing. In addition to previously described modality-specific differences, the present study showed that the left calcarine gyrus and the right caudate were additionally recruited in deaf compared with hearing individuals. In addition, this study showed that the bilateral posterior superior temporal gyrus is shaped by cross-modal plasticity, whereas the left frontotemporal areas are shaped by early language experience. Although an overall left-lateralized pattern for language processing was observed in the early deaf individuals, regional lateralization was altered in the inferior temporal gyrus and anterior temporal lobe. These findings indicate that the core language network functions in a modality-independent manner, and provide a foundation for determining the contributions of sensory and linguistic experiences in shaping the neural bases of language processing.
This study examined the language outcomes of children with mild to severe hearing loss during the preschool years. The longitudinal design was leveraged to test whether language growth trajectories ...were associated with degree of hearing loss and whether aided hearing influenced language growth in a systematic manner. The study also explored the influence of the timing of hearing aid fitting and extent of use on children's language growth. Finally, the study tested the hypothesis that morphosyntax may be at particular risk due to the demands it places on the processing of fine details in the linguistic input.
The full cohort of children in this study comprised 290 children who were hard of hearing (CHH) and 112 children with normal hearing who participated in the Outcomes of Children with Hearing Loss (OCHL) study between the ages of 2 and 6 years. CHH had a mean better-ear pure-tone average of 47.66 dB HL (SD = 13.35). All children received a comprehensive battery of language measures at annual intervals, including standardized tests, parent-report measures, and spontaneous and elicited language samples. Principal components analysis supported the use of a single composite language score for each of the age levels (2, 3, 4, 5, and 6 years). Measures of unaided (better-ear pure-tone average, speech intelligibility index) and aided (residualized speech intelligibility index) hearing were collected, along with parent-report measures of daily hearing aid use time. Mixed modeling procedures were applied to examine the rate of change (227 CHH; 94 children with normal hearing) in language ability over time in relation to (1) degree of hearing loss, (2) aided hearing, (3) age of hearing aid fit and duration of use, and (4) daily hearing aid use. Principal components analysis was also employed to examine factor loadings from spontaneous language samples and to test their correspondence with standardized measures. Multiple regression analysis was used to test for differential effects of hearing loss on morphosyntax and lexical development.
Children with mild to severe hearing loss, on average, showed depressed language levels compared with peers with normal hearing who were matched on age and socioeconomic status. The degree to which CHH fell behind increased with greater severity of hearing loss. The amount of improved audibility with hearing aids was associated with differential rates of language growth; better audibility was associated with faster rates of language growth in the preschool years. Children fit early with hearing aids had better early language achievement than children fit later. However, children who were fit after 18 months of age improved in their language abilities as a function of the duration of hearing aid use. These results suggest that the language learning system remains open to experience provided by improved access to linguistic input. Performance in the domain of morphosyntax was found to be more delayed in CHH than their semantic abilities.
The data obtained in this study largely support the predictions, suggesting that mild to severe hearing loss places children at risk for delays in language development. Risks are moderated by the provision of early and consistent access to well-fit hearing aids that provide optimized audibility.
In this modern era, language has no geographic boundary. Therefore, for developing an automated system for search engines using audio, tele-medicine, emergency service via phone etc., the first and ...foremost requirement is to identify the language. The fundamental difficulty of automatic speech recognition is that the speech signals vary significantly due to different speakers, speech variation, language variation, age and sex wise voice modulation variation, contents and acoustic conditions and so on. In this paper, we have proposed a deep learning based ensemble architecture, called FuzzyGCP, for spoken language identification from speech signals. This architecture combines the classification principles of a Deep Dumb Multi Layer Perceptron (DDMLP), Deep Convolutional Neural Network (DCNN) and Semi-supervised Generative Adversarial Network (SSGAN) to increase the precision to maximum and finally applies Ensemble learning using Choquet integral to predict the final output, i.e., the language class. We have evaluated our model on four standard benchmark datasets comprising of two Indic language datasets and two foreign language datasets. Irrespective of the languages, the F1-score of the proposed language identification model is as high as 98% in MaSS dataset and worst performance is that of 67% on the VoxForge dataset which is much better compared to maximum of 44% by state-of-the-art models on multi-class classification. The link to the source code of our model is available here.
•Developed an automatic spoken language identification system called FuzzyGCP.•Deep Dumb MLP is used as a classifier for numeric features.•Image based features are fed to Deep CNN and Semi-supervised GAN models.•Formed an ensemble by uniting results of 3 models using a fuzzy integral measure.•Evaluated the model on 4 standard publicly available multi-lingual datasets.
PDF
