Accueil > Actualités

Séminaires LPNC

publié le , mis à jour le



  • Automated text simplification as a reading aid for low-vision individuals
    16 mars 2021 - 13:00 - Aurélie Calabrese - Inria, Sophia-Antipolis

    Résumé : In developed countries, the majority of people with visual impairment are legally blind, but not totally blind. Instead, they have what is referred to as low vision, commonly caused by Central visual Field Loss (CFL). This degenerative condition is caused by non-curable retinal diseases, such as Age-related Macular Degeneration (DMLA in French). Patients suffering from such pathologies will develop a blind region called scotoma, located at the center of their visual field and spanning about 20° or more. To better visualize the impact of such a large hole in your visual field, try stretching your index and little finger as far as possible from each other at arm’s length ; the span is about 15°. Central vision cannot be restored and difficulty with reading becomes the primary complaint of patients seeking rehabilitation. To help CFL individuals improve their reading performance, it is necessary to investigate the underlying causes of their deficit, to then overcome them with specific adjustments.
    In this presentation, I propose to address the issue of reading with CFL from a new perspective, a linguistic one, that takes into account the whole complexity of texts. I will present a series of experiments that investigate what makes a text especially complex when reading with CFL, by assessing the effect of lexical properties on reading speed. We will focus on three word properties : frequency, orthographic similarity and length. I will conclude on the relevance of this work to design text simplification tools, customised to the specific needs of readers with CFL, to be used as efficient reading aids for this population.

    Lieu : Salle virtuelle Zoom du laboratoire (mot de passe : cognition)

  • Automatic imitation of speech : mechanisms governing perception-production links
    13 avril 2021 - 13:00 - Patti Adank - University College, Londres (UCL)

    Résumé : We automatically and spontaneously imitate other people’s speech and other actions in daily interaction. An extensive neuroscience literature has demonstrated that observing people speak activates neural substrates associated with speech production, thus demonstrating the tight link between speech perception and production, as also evident from neuroimaging results. We therefore automatically (and usually covertly) show an imitative response when observing other people’s speech. The link between a sensory stimulus and its associated response can be investigated using the Stimulus Response Compatibility (SRC) paradigm. It is assumed that the automatic imitative response upon action observation is due to overlapping representations between a stimulus and its associated response. This paradigm has been used in the last 20 years to study whether and how observed actions and their responses are linked to the action production system. In this talk, I will outline how we have applied the SRC paradigm to imitation of speech, with the ultimate aim of elucidating the conditions under which we imitate speech, as well as clarifying how speech production links to speech perception. I will discuss a series of behavioural studies using the SRC paradigm, both lab-based and online studies, that investigated if we in fact really ’automatically’ imitate speech stimuli, how familiarity with the speech stimulus affects automatic imitation, and whether and how perception-production links for speech can be modified using training.

    Lieu : Salle virtuelle Zoom du laboratoire (mot de passe : cognition)

  • Le regard dans la cognition humaine : de la recherche fondamentale à la recherche translationnelle
    27 avril 2021 - 13:00 - Laurence Conty - Laboratoire CHArt, Université Paris-Nanterre

    Résumé : Un nombre important de données empiriques étaye aujourd’hui l’hypothèse que le regard d’autrui module automatiquement différents processus cognitifs concomitants ou subséquents au regard. Les recherches focalisent généralement leur intérêt soit sur les effets induits par regard direct (i.e. qui crée un contact visuel avec le participant) ou ceux induits par le regard dévié (i.e. détourné vers l’espace environnant). Dans un premier temps, je développerai l’argument que cette dichotomie a une valeur plus pragmatique que théorique en questionnant notamment les mécanismes cognitifs et cérébraux sous-tendant le traitement du regard. Je focaliserai ensuite mon intervention sur trois effets reproductibles : l’amélioration de l’intéroception et de la mémoire des visages induit par le regard direct, et amélioration de l’orientation attentionnelle induit par le regard dévié. Je soulignerai que ces effets sont notablement bénéfiques et présenterai les nouveaux projets de mon équipe qui cherchent à déterminer, dans une démarche translationnelle, leur potentiel thérapeutique ou diagnostic, par exemple dans la maladie d’Alzheimer ou la Négligence Spatiale Unilatérale (NSU).

    Lieu : Salle virtuelle Zoom du laboratoire (mot de passe : cognition)

  • On the measurement and estimation of cognitive processes with electrophysiological recordings and reaction time modeling
    25 mai 2021 - 13:00 - Gabriel Weindel - Laboratoire de Neurosciences Cognitives, Aix-Marseille et Laboratoire de Psychologie Cognitive, Aix-Marseille

    Résumé : Mathematical models of decision making aim at decomposing reaction times into processing units by formalizing an assumed generative model. Unfortunately, a given generative model may explain the behavioral data while not necessarily reflecting the underlying cognitive processes. Obtaining measurements between the stimulus and the responses could provide additional information that fruitfully constrains the processing assumptions. In this spirit, we used electrophysiological recordings (electromyography and electroencephalography) to decompose each reaction time into different intervals, presumed to contain the different processing units assumed in a model. Simultaneously, we manipulated time-honored experimental factors to compare the cognitive locus of experimental effects inferred from either electrophysiological recordings or model fitting procedures.
    We show that the inferences drawn from cognitive mathematical models conflict with the electrophysiological decomposition when : 1) the model’s core assumption of independence between decision and non-decision processes is proven to be false ; 2) standard modeling strategies are inadequate to capture the locus of an experimental effect revealed by the electrophysiological decomposition ; and 3) opposite experimental effects are revealed across processing units. These results constitute a thorough examination of the validity of a popular decision making model. They offer new insights on the information processes that allow humans to decide between alternatives.

Ajouter un événement iCal