Modèles computationnels du développement cognitif

Leader

Research center

45 rue d’Ulm
75230 Paris
Marc Mézard

Institution

Ecole Normale Supérieure
CNRS
Ecole des Hautes Etudes en Sciences Sociales
ED 158

Laboratory

Laboratoire de Sciences Cognitives et Psycholinguistique
UMR 8554
Labex IEC

Mots clefs

Neurosciences computationnelles
Psycholinguistique
acquisition du langage
machine learning
traitement du signal
 

publications

Ludusan B, Cristia A, Martin A, Mazuka R, Dupoux E. Learnability of prosodic boundaries: Is infant-directed speech easier? J Acoust Soc Am. 2016 Aug;140(2):1239. doi: 10.1121/1.4960576.

Buon M, Jacob P, Margules S, Brunet I, Dutat M, Cabrol D, Dupoux E. Friend or foe? Early social evaluation of human interactions. PLoS One. 2014 Feb 19;9(2):e88612. doi: 10.1371/journal.pone.0088612. eCollection 2014.

Ngon, Martin, #Dupoux, et al .(2013). Nonwords, nonwords, nonwords. Developmental Science, 16(1), 24-34.

Minagawa-Kawai, ... & #Dupoux (2013). Insights on NIRS sensitivity. Frontiers in Language Sciences, 4(170.

Martin , Peperkamp, & #Dupoux (2013). Learning Phonemes with a Proto-lexicon. Cognitive Science, 37, 103-124.

Jansen, #Dupoux, et al. (2013). Zero resource speech technologies and models of early language acquisition. Proceedings of ICASSP 2013.

Cristia, #Dupoux, et al. (2013). An online database of infant functional Near InfraRed Spectroscopy. PLoS One, 8(3), e58906.

Fields of research

Cognitive neurosciences / neuropsychology /neuroeconomy

Research Theme

During the last thirty years, developmental psychology have documented the surprising speed and robustness with which babies learn the linguistics and social characteristics of their maternal environment. During the first year alone, infants achieve impressive landmarks regarding three key language components (see Figure 1). First, they tune into the phonemic categories (consonants and vowels) of their language, i.e., they lose the ability to distinguish some fine phonetic contrasts that belong to the same category, enhance their ability to distinguish some between-category contrasts, and refine their ability to ignore acoustic variations due to speaker characteristics. Second, infants refine their ability to segment the speech stream, from large prosodic units to small ones, and determine the suprasegmental features that are relevant for word recognition. Third, infants start to extract very frequent word patterns from their environnment, and compile segmentation strategies that help them constructing a recognition lexicon. These aspects of language are learned efforlessly even through they are difficult for adults to learn later in life. Moroever, they are learned without direct parental supervision. Finally, and most suprisingly, the acquisition of these three components is not done sequentially, but in a largely overlapping fashion.

 

Etudiants ENP

Maria Julia CARBAJAL

Membres de l'équipe

LUDUSAN Bogdan
DUNBAR EWAN
SYNNAEVE Gabriel
CAO Xuan-Nga

Lab rotation

Modeling early language

Chercheur responsable: 

DUPOUX Emmanuel

Dates: 

2 January 2018 - 29 June 2018

Date limite de candidature: 

29 June 2018

Period

~ Jan-March 2018

~ April-June 2018

Project

Recently, machine learning has achieved spectacular results in reaching human-level performance on a variety of tasks (vision, language processing). Can these systems be used as quantitative models of brain processes? The internship will explore this question on speech or language as observed in infants or adults.

Some familiarity with a programming language and data analysis is a prerequisite. www.syntheticlearner.net

Contact

Ecole Normale Supérieure - Laboratoire de Sciences Cognitives et Psycholinguistique - 29, rue d'Ulm 75006 Paris - +33 1 44 32 26 17

Superviseur: 

DUPOUX Emmanuel