Talks & Courses

Sparse coding is a technique used in signal processing and machine learning to represent data in a more concise and efficient manner. …

Les aires visuelles doivent transformer le signal lumineux brut en une représentation qui transmet efficacement des informations sur …

Related papers Jens Kremkow, Laurent U Perrinet, Cyril Monier, Jose-Manuel Alonso, Ad M Aertsen, Yves Frégnac, Guillaume S Masson (2016). Push-Pull Receptive Field Organization and Synaptic Depression: Mechanisms for Reliably Encoding Naturalistic Stimuli in V1.

The biological retina represents changes in light intensity and communicates them to the brain by means of spikes. Similarly, event-based cameras mimic the way the biological retina processes visual information: each pixel independently reports brightness changes as asynchronous temporal events. This differs from conventional cameras, which capture static images at fixed intervals. This event-based paradigm enables novel visual computations with high speed, low latency, and energy efficiency. I will highlight recent work applying event cameras to tasks such as visual odometry, motion detection, or gesture recognition. The goal is to demonstrate the benefits for computer vision of emulating biological principles inspired by neuroscience, but also to deepen our understanding of visual processes.

Event-based cameras mimic the way biological retinas process visual information: each pixel independently reports brightness changes as asynchronous temporal events. This departs from conventional cameras that capture static frames at fixed intervals. I will first discuss how the biological retina detects light intensity changes and communicates this to the brain. Compared to traditional cameras, the event-based paradigm enables new vision applications with high-speed, low latency and energy-efficiency. I will highlight recent works applying event cameras to tasks such as visual odometry, motion detection or gesture recognition. The goal is to demonstrate the advantages for computer vision that emulate biological principles inspired by neurosciences.

Event-based cameras mimic the way biological retinas process visual information: each pixel independently reports brightness changes as asynchronous temporal events. This departs from conventional cameras that capture static frames at fixed intervals. I will first discuss how the biological retina detects light intensity changes and communicates this to the brain. Compared to traditional cameras, the event-based paradigm enables new vision applications with high-speed, low latency and energy-efficiency. I will highlight recent works applying event cameras to tasks such as visual odometry, motion detection or gesture recognition. The goal is to demonstrate the advantages for computer vision that emulate biological principles inspired by neurosciences.

Recently, there has been growing interest in exploring the hypothesis that neural activity conveys information through precise spiking motifs. To investigate this hypothesis, various algorithms have been proposed to detect such motifs in spiking activity recorded from populations of neurons. In this study, we present a detection model that takes the form of logistic regression combined with temporal convolution. A key advantage of this model is its differentiability, which allows us to formulate a gradient descent on any appropriate loss. We prove its efficiency on synthetic data for which the ground truth is available to use supervised learning. However, this ground truth information is not available for neurobiological data, where a self-learning procedure is required. We show that a contrastive learning method can recover the synthetically generated spiking motifs without knowing the ground truth. In the future, we aim to extend this method to real neurobiological data to explore and detect spiking motifs in a more natural and biologically relevant context.

Hybrid Session, Room 2 Chair: Sander Bohté, Sebastian Otte read the proceedings The code is available on GitHub

Event-based cameras mimic the way biological retinas process visual information: each pixel independently reports brightness changes as asynchronous temporal events. This departs from conventional cameras that capture static frames at fixed intervals. I will first discuss how the retina detects light intensity changes and communicates this to the brain. Compared to traditional cameras, the event-based paradigm enables new vision applications with high-speed, low latency and energy-efficiency. I will highlight recent works applying event cameras to tasks such as visual odometry, motion detection or gesture recognition. The goal is to demonstrate the advantages for computer vision that emulate biological principles.

Visual areas are essential in transforming the raw luminous signal into a representation which efficiently conveys information about …

Visual areas are essential in transforming the raw luminous signal into a representation which efficiently conveys information about …

workshop organisé par les étudiants du master de sciences cognitives les 23 et 24 janvier 2023.

During the #CVPR2022-#NeuroVision workshop about « What can computer vision learn from visual neuroscience? », @A_Grismaldi will talk today about « Learning hetero-synaptic delays of Spiking Neurons for motion detection » shows how to learn spike motifs !

Follows a previous work Emmanuel Dauce, Pierre Albigès, Laurent U Perrinet (2020). A dual foveal-peripheral visual processing model implements efficient saccade selection. Journal of Vision. Preprint PDF Cite Code DOI for a follow-up, check out Jean-Nicolas Jeremie, Emmanuel Dauce, Laurent U Perrinet (2022).

for visual search see: Emmanuel Dauce, Pierre Albigès, Laurent U Perrinet (2020). A dual foveal-peripheral visual processing model implements efficient saccade selection. Journal of Vision. Preprint PDF Cite Code DOI for retinotopy, see: Jean-Nicolas Jeremie, Emmanuel Dauce, Laurent U Perrinet (2022).

Follow this future presentations Antoine Grimaldi, Laurent U Perrinet (2022). Learning heterogeneous delays of Spiking Neurons for motion detection. NeuroVision Workshop in conjunction with CVPR 2022. PDF Cite URL Glad to meet the @centuri_ls crowd at the #CENTURIday !

Où: Salle PHY51 - Marseille (France) Quoi: Master 1 Neurosciences et Sciences Cognitives Réseaux neuronaux artificiels pour la vision Mercredi 23/03/2022 de 9h-12h Introduction aux Neurosciences de la Vision Réseaux de neurones artificiels et apprentissage machine slides Neurones impulsionnels et modèles des fonctions visuelles Mercredi 23/03/2022 de 13h30-16h30 TP via notebook https://github.

In most mammals, the primary visual cortex (V1) processes complex mixtures of orientations to construct an accurate neural …

Les illusions visuelles sont des créations d’artistes, de scientifiques et plus récemment, grâce aux réseaux sociaux, du grand …

The primary visual cortex (V1) processes complex mixtures of orientations to build neural representations of our visual environment. It …

Presenting my poster tonight at 8:00p #cosyne2020, a work developed using Sparse Deep Predictive Coding (SDPC) during my PhD @laurentperrinet @NeuroSchool_mrs pic.twitter.com/LtUEBnlPNt — AF (@Angelo_RDN) February 28, 2020 This is from:

The primary visual cortex (V1) processes complex mixtures of orientations to build neural representations of our visual environment. It …

Building models which efficiently process images is a great source of inspiration to better understand the processes which underly our …

Visual search is an essential cognitive ability, offering a prototypical control problem to be addressed with Active Inference. Under a …

Visual areas are essential in transforming the raw luminous signal into a representation which efficiently conveys information about …

Les illusions visuelles sont des créations d’artistes, de scientifiques et plus récemment, grâce aux réseaux sociaux, du grand …

In computer vision, the visual search task consists in extracting a scarce and specific visual information (the target) from a large …

Animal behavior must constantly adapt to changes, for example when the state of an environmental context changes unexpectedly. For an …

Les objectifs sont : – mieux comprendre la fonction de la perception visuelle en explorant certaines limites ; – mieux comprendre l’importance de l’aspect dynamique de la perception ; – mieux comprendre le rôle de l’action dans la perception.

Animal behavior has to constantly adapt to changes, for instance when unexpectedly switching the state of an environmental context. For …

Visual areas are essential in transforming the raw luminous signal into a representation which efficiently conveys information about …

Visual areas are essential in transforming the raw luminous signal into a representation which efficiently conveys information about …

See the final publication @ Chloe Pasturel, Anna Montagnini, Laurent U Perrinet (2020). Humans adapt their anticipatory eye movements to the volatility of visual motion properties. PLoS Computational Biology. Preprint PDF Cite Code DOI previous talk @ LAW, Lyon previous talk @ INVIBE FEST, Paris previous talk @ Brain workshop, Marseille next talk @ CAUSAL Kick-off, Marseille

Le jeudi 10 janvier 2019, je suis venu échanger au côté de Serge Dentin autour de films traitant du rapport fiction/réel, des illusion …

in “La modélisation de la genèse physico-mathématique du vivant” BIOMORPHISME ET CREATION ARTISTIQUE – Session 3 Date 11 Octobre 2018 Atelier Séminaire/workshop organisé dans le cadre du projet Biomorphisme. Approches sensibles et conceptuelles des formes du vivant http://lesa.

FÊTE DE LA SCIENCE 2018 : Alcazar / MERLAN L’Association Polly Maggoo http://www.pollymaggoo.org/ met en place tout le long de l’année, des actions de culture scientifique et artistique en direction du grand public et des lycées, au cours desquelles l’association programme des films à caractère scientifique.

See the final publication @ Chloe Pasturel, Anna Montagnini, Laurent U Perrinet (2020). Humans adapt their anticipatory eye movements to the volatility of visual motion properties. PLoS Computational Biology. Preprint PDF Cite Code DOI previous talk @ LAW, Lyon previous talk @ INVIBE FEST, Paris next talk @ LACONEU, Chile next talk @ CAUSAL Kick-off, Marseille next talk @ NeuroFrance, Marseille

See the final publication @ Chloe Pasturel, Anna Montagnini, Laurent U Perrinet (2020). Humans adapt their anticipatory eye movements to the volatility of visual motion properties. PLoS Computational Biology. Preprint PDF Cite Code DOI previous talk @ LAW, Lyon next talk @ Brain workshop, Marseille next talk @ LACONEU, Chile next talk @ CAUSAL Kick-off, Marseille next talk @ NeuroFrance, Marseille

Meetup Art et Neurosciences Quoi Meetup Art et Neurosciences Qui Association NeuroNautes Quand 25 Janvier 2018 Où Salle des voutes campus Saint Charles Support visuel https://laurentperrinet.github.io/sciblog/files/2018-01-25_meetup-neuronautes.html (notes: la présentation peut mettre un certain temps à charger.

see a related work describing SDPC in: Victor Boutin, Angelo Franciosini, Frederic Chavane, Franck Ruffier, Laurent U Perrinet (2021). Sparse Deep Predictive Coding captures contour integration capabilities of the early visual system.

FESTIVAL INTERFÉRENCES​ Cinéma Documentaire et Débat Public FESTIVAL INTERFÉRENCES​ Le collectif Scènes Publiques composé de citoyens, chercheurs et cinéastes, organise la deuxième édition du Festival Interférences du 8 au 18 novembre 2017 à Lyon.

This seminar is an exercise to introduce the AMU masters into the format of international conferences. As such, we will try to …

RENCONTRES INTERNATIONALES SCIENCES & CINÉMAS cinéma les Variétés http://pollymaggoo.org/wp-content/uploads/2016/10/RISC2016_A3-724x1024.jpg L’Association Polly Maggoo http://www.pollymaggoo.org/ programme la 10e édition des RENCONTRES INTERNATIONALES SCIENCES & CINÉMAS (RISC) à Marseille, au cours desquelles l’association programme des films à caractère scientifique.

Based on Perrinet et al, 2012 and Khoei et al, 2013 See a followup in Khoei et al, 2017

Natural environments potentially contain several interesting targets for goal-directed behavior. Thus sensorimotor systems need to …

See a followup in Laurent U Perrinet, James A Bednar (2015). Edge co-occurrences can account for rapid categorization of natural versus animal images. Scientific Reports. Preprint PDF Cite Code DOI

See the final publication @ Chloe Pasturel, Anna Montagnini, Laurent U Perrinet (2020). Humans adapt their anticipatory eye movements to the volatility of visual motion properties. PLoS Computational Biology. Preprint PDF Cite Code DOI next talk @ INVIBE FEST, Paris next talk @ Brain workshop, Marseille next talk @ LACONEU, Chile next talk @ CAUSAL Kick-off, Marseille next talk @ NeuroFrance, Marseille

See a followup in Perrinet et al, 2012

Les illusions visuelles sont des créations d’artistes, de scientifiques et plus récemment, grâce aux réseaux sociaux, du grand …

Ce lundi 25 avril de 9h à 12h, je suis venu échanger au côté de Serge Dentin autour de films traitant du rapport fiction/réel, des …

We stand at a point in history where our phones have become smart but lack a feature which prevails in most forms of living …

We stand at a point in history where our phones have become smart but lack a feature which prevails in most forms of living …

see Kaplan and al, 2014

The question how the visual system is able to create a coherent representation of a rapidly changing environment in the presence of …

Moving objects generate sensory information that may be noisy and ambiguous, yet it is important to be able to reconstruct object speed …

Together with Bernhard Kaplan, we talked about how we aim at “compiling” a predictive motion-based approach as a spiking neural networks and then as a parallel wafer systems in the BrainscaleS project (Demo 1, Task4).

See a followup in Laurent U Perrinet, James A Bednar (2015). Edge co-occurrences can account for rapid categorization of natural versus animal images. Scientific Reports. Preprint PDF Cite Code DOI

This session aims at presenting new ideas that emerged during the first years of BrainScaleS. Indeed, the collaborations that were …

Oriented edges in images of natural scenes tend to be aligned in collinear or co-circular arrangements, with lines and smooth curves …

Moving objects generate sensory information that may be noisy and ambiguous, yet it is important to be able to reconstruct object speed …

Oriented edges in images of natural scenes tend to be aligned in collinear or co-circular arrangements, with lines and smooth curves …

In low-level sensory systems, it is still unclear how the noisy information collected locally by neurons may give rise to a coherent …

Oriented edges in images of natural scenes tend to be aligned in collinear or co-circular arrangements, with lines and smooth curves …

Oriented edges in images of natural scenes tend to be aligned in collinear or co-circular arrangements, with lines and smooth curves …

Sensory informations such as visual images are inherently variable. We use probabilistic models to describe how the low-level visual …

Sensory informations such as visual images are inherently variable. We use probabilistic models to describe how the low-level visual …

En perception, les neurones « parlent » tous en même temps par de brèves impulsions électrochimiques, générant un mélange de signaux, …

see this more recent talk @ UCL, London

Short presentation of a large moving pattern elicits an ocular following response that exhibits many of the properties attributed to …

Nous parlerons de cette partie "mécanique" du cerveau animal ou humain qui permet de percevoir les mouvements et de … …

see this subsequent paper in the Journal of Computational Neuroscience

Moving the eyes rapidly to track a visual object moving in a cluttered environment is an essential function. However, doing so rapidly …

The machinery behind the visual perception of motion and the subsequent sensorimotor transformation, such as in Ocular Following …

Computational Neuroscience is a synthetic, inter-disciplinary approach aiming at understanding cognition by analyzing the mechanisms …

related publications @ FENS 2006, @ NeuroComp 2008 and @ AREADNE 2008

I will illustrate in this talk how computational neuroscience may inspire and be inspired by mathematical image processing. Focusing on …

The quality of the representation of an object’s motion is limited by the noise in the sensory input as well as by an intrinsic …