Bayesian-Modelling | Laurent Perrinet

[Lennie, 2003, The Cost of Cortical Computation]

Neurosciences and sparsity: a survey

Go to wooclap.com
Enter the code HLEQUP
Or directly follow https://app.wooclap.com/HLEQUP?from=instruction-slide

Neurosciences and sparsity: a survey

Neurosciences and sparsity

Neurosciences and sparsity

[Brunel, 2001]

Neurosciences and sparsity

[Mainen & Sejnowski, 1995]

Neurosciences and sparsity

mainen et sejnowski diesmann

Neurosciences and sparsity

Sparse representations in a nutshell

in summary: Sparse representations resulting from these processes have been successfully applied in various domains such as image processing, computer vision, and audio signal processing. It has shown promise in tasks such as noise reduction, compression, feature extraction, and pattern recognition. By capturing the essential structure and characteristics of the data in a sparse representation, sparse coding can help reduce redundancy and noise, and extract meaningful features for further analysis or processing.

…let’s delve into a computational theory of sparse coding

review_bib = s.content_bib(“LP”, “2015”, ‘“Sparse models” in Biologically Inspired Computer Vision’)

Sparse representations in a nutshell

Sparse representations in a nutshell

Sparse representations in a nutshell

Generative model of image synthesis:

$I[x, y] = $ $\sum_{i=1}^{K} a[i] \cdot \phi[i, x, y]$ $ + \varepsilon[x, y]$ Where $\phi$ is a dictionary of $K$ atoms, $a$ is a sparse vector of coefficients, and $\varepsilon$ is a noise term.

Sparse representations in a nutshell

Sparse representations in a nutshell

Given an observation $I$,

$$ \begin{aligned} \mathcal{L}(a) & = - \log Pr( a | I ) \\ \end{aligned} $$

Sparse representations in a nutshell

Given an observation $I$,

$$ \begin{aligned} \mathcal{L}(a) & = - \log Pr( a | I ) \\ & = - \log Pr( I | a ) - \log Pr(a) \\ \end{aligned} $$

Sparse representations in a nutshell

Given an observation $I$,

$$ \begin{aligned} \mathcal{L}(a) & = - \log Pr( a | I ) \\ & = - \log Pr( I | a ) - \log Pr(a) \\ & = \frac{1}{2\sigma_n^2} \sum_{x, y} ( I[x, y] - \sum_{i=1}^{K} a[i] \cdot \phi[i, x, y])^2 - \sum_{i=1}^{K} \log Pr( a[i] ) \end{aligned} $$

Sparse representations in a nutshell

The problem is formalized as an optimization problem $a^\ast = \arg \min_a \mathcal{L}(a)$ with:

$$ \mathcal{L} = \frac{1}{2} \sum_{x, y} ( I[x, y] - \sum_{i=1}^{K} a[i] \cdot \phi[i, x, y])^2 + \lambda \cdot \sum_i ( a[i] \neq 0) $$

Sparse representations in a nutshell

The problem is formalized as an optimization problem $a^\ast = \arg \min_a \mathcal{L}(a)$ with:

$$ \mathcal{L}(a) = \frac{1}{2} \sum_{x, y} ( I[x, y] - \sum_{i=1}^{K} a[i] \cdot \phi[i, x, y])^2 + \lambda \cdot \sum_{i=1}^{K} | a[i] | $$

Sparse representations in a nutshell

[Rentzeperis et al (2023)]

---

## Sparse representations and learning

Convolutional Sparse Coding

Convolutional Sparse Coding

Jean-Nicolas Jérémie (2025). Soutenance de Jean-Nicolas Jérémie "Foveal Retinotopy and Dual Pathways: A Computational Model for Active Visual Search".

novel challenges for CNNs

1/ backpropagation is not bioplausible

CNN: Predictive processing

result= interpretable features

SNN in neuromorphic engineering

The HD-SNN neural network.

2 MINUTE

For instance, we show how precise spike times may be used to detect the direction of motion from such a stream of events in an ultrafast fashion.

SNN in neuromorphic engineering

2 MINUTE

A nice HSD neuron

For instance, we show how precise spike times may be used to detect the direction of motion from such a stream of events in an ultrafast fashion.

SNN in neuromorphic engineering

The HD-SNN neural network.

2 MINUTE

nice kernels

SNN in neuromorphic engineering

The HD-SNN neural network.

2 MINUTE

frugal computing

Neurosciences and sparsity

Laurent Perrinet

Séminaire à l’atelier “IA embarquée” du PEPR IA

[2026-01-29]

Contact me @ laurent.perrinet@univ-amu.fr

Foveal Retinotopy and Dual Pathways: A Computational Model for Active Visual Search

Fri, 10 Oct 2025 00:00:00 +0000

Abstract

This thesis investigates visual search through the lens of the dual visual pathways found in biological systems : the ventral (“what”) pathway, involved in object recognition, and the dorsal (“where”) pathway, responsible for spatial localisation and saccadic planning. Drawing from both neuroscience and computer vision, we propose a computational framework that integrates deep convolutional neural networks (DCNNs) within a biologically inspired architecture grounded in foveal retinotopy. As a proof of concept, prior work has demonstrated that incorporating saccadic planning improves digit categorisation performance in a controlled environment. Building upon this foundation, the primary objective of this thesis is to extend the computational framework to natural images in more ecologically valid settings. Our contributions are as follows : (1) We introduce a novel framework for training and evaluating DCNNs using semantically grounded, task-specific labels ; (2) We bridge the gap between artificial models and biological substrates by emphasizing the role of foveal retinotopy in robust object categorisation and precise localisation ; (3) We disentangle the interplay between categorisation and localisation by proposing a novel “localisation-frame” dataset, aimed at guiding the design of a biologically plausible dorsal stream model ; and (4) We present an initial model of the dorsal pathway, leveraging the new dataset to develop interpretable and efficient active vision systems—where interpretability is achieved through modular and spatially structured representations, and efficiency is reflected in reduced computational cost during inference with saccade planning. Overall, this thesis extends the dual-stream computational paradigm for visual search, contributes tools for explainable active vision, and offers a platform to explore hypotheses about functional specialisation in the human visual cortex.

Keywords

Visual search, Dual visual pathways, Deep Convolutional Neuronal, Network, Foveal retinotopy, Active vision

Résumé

Cette thèse étudie la recherche visuelle à travers le prisme des deux voies visuelles identifiées dans les systèmes biologiques : la voie ventrale, impliquée dans la reconnaissance des objets, et la voie dorsale, responsable de la localisation spatiale et de la planification des saccades. S’inspirant à la fois des neurosciences et de la vision artificielle, nous proposons un cadre computationnel intégrant des réseaux neuronal convolutifs profonds (DCNN) dans une architecture biologiquement plausible, fondée sur la rétinotopie fovéale. Des travaux antérieurs ont démontré que l’intégration de la planification des saccades améliorait les performances de catégorisation de chiffres dans un environnement contrôlé. S’appuyant sur cette base, l’objectif principal de cette thèse est d’étendre ce cadre théorique à des images naturelles dans des contextes plus écologiquement valides. Nos contributions sont les suivantes : (1) Nous proposons un nouveau cadre de travail pour l’entraînement et l’évaluation des DCNN, basé sur la sémantique sous-jacente aux labels initialement définis dans la communauté de la recherche computationnelle, ce qui permet de définir des tâches écologiques spécifiques ; (2) nous rapprochons les modèles artificiels des substrats biologiques en soulignant le rôle crucial de la retinotopie fovéales pour une catégorisation robuste et une localisation précise. (3) Nous approfondissons la connaissance de l’interaction entre la catégorisation et la localisation en proposant un ensemble de résultats structuré autour de cette relation, afin de guider la conception d’un modèle plausible de la voie dorsale ; (4) Enfin, en nous appuyant sur ces résultats, nous proposons une première modélisation de la voie dorsale visant à développer des systèmes de vision active à la fois interprétables, grâce à des représentations modulables et spatialement structurées, et efficaces, grâce à la planification de saccades permettant de réduire les coûts de calcul liés à l’inférence. Dans l’ensemble, cette thèse apporte plusieurs éléments : elle enrichit le modèle de vision artificielle des deux voies majeures impliquées dans la recherche visuelle, elle permet de développer des outils de vision active interprétables et elle fournit un cadre pour étudier les hypothèses biologiques relatives à la spécialisation fonctionnelle des aires cérébrales dédiées à la vision chez l’être humain.

Mots-clés

Recherche visuelle, Voie visuel ventrale, Voie visuel dorsale, Réseau neuronal convolutifs profonds, Rétinotopie fovéale, Vision active

Soutenance de Jean-Nicolas Jérémie "Foveal Retinotopy and Dual Pathways: A Computational Model for Active Visual Search"

Fri, 10 Oct 2025 00:00:00 +0000

Jean-Nicolas Jérémie soutiendra publiquement ses travaux de thèse intitulés: Foveal Retinotopy and Dual Pathways: A Computational Model for Active Visual Search

dirigés par Monsieur Laurent PERRINET et Monsieur Emmanuel DAUCE

Date: le vendredi 10 octobre 2025 à 13h30
Lieu : Faculté de Médecine de la Timone 27 Boulevard Jean Moulin, 13005 Marseille 5ème
Salle : Amphithéâtre CERIMED

Jean-Nicolas Jérémie (2025). Foveal Retinotopy and Dual Pathways: A Computational Model for Active Visual Search.

Cite

Composition du jury proposé

Name	Affiliation	Role
M. Ronan SICRE	IRIT (UMR 5505) – Université de Toulouse III	Rapporteur
M. Jean‑Julien AUCOUTURIER	FEMTO‑ST (UMR 6174) – Université de Bourgogne Franche‑Comté	Rapporteur
Mme Teresa SERRANO‑GOTARREDONA	IMSE‑CNM‑CSIC – Universidad de Sevilla	Examinatrice
M. Franck RUFFIER	Lab‑STICC (UMR 6285) – ENSTA\|IP Paris	Examinateur
M. Matthieu GILSON	INT (UMR 7289) – Aix Marseille Université	Président
M. Laurent PERRINET	INT (UMR 7289) – Aix Marseille Université	Directeur de thèse
M. Emmanuel DAUCÉ	Centrale Méditerranée	Co‑directeur de thèse

Résumé

Abstract

Artificial neural networks and machine learning applied to the understanding of biological vision

Mon, 26 May 2025 14:00:00 +0000

Master M4NC de l’institut NeuroMod, cours Prospective Innovation and Research.

2025-03-11-phd-program-sparse-representations

Tue, 11 Mar 2025 00:00:00 +0000

Sparse representations

Laurent Perrinet

NeuroSchool PhD Program in Neuroscience

[2025-03-11]

Code / Contact me @ laurent.perrinet@univ-amu.fr

Sparse representations?

Paysage catalan (Le Chasseur)

to rephrase the expression “The Unreasonable Effectiveness of Mathematics” by Wigner, the “Unreasonable efficiency of vision” is playfully illustrated in this painting from Joan Miró, which allows us to depict this Catalan landscape with the a few strokes where our imagination will fill the gaps and signify the landscape, allowing us to imagine the hunter, the sardine or the plane.

the whole is the sum of a few parts

Sparse coding is a technique used in signal processing and machine learning to represent data in a more concise and efficient manner. It aims to find a sparse representation of the data, which means representing the data with only a small number of non-zero coefficients or activations. In sparse coding, a set of basis functions or atoms is typically defined, and the goal is to find a linear combination of these atoms that best represents the input data. The coefficients of this linear combination are often constrained to be sparse, meaning that only a few of them are allowed to be non-zero.

Sparse representations in computer vision

Sparse representations in computer vision

Sparse representations in computer vision

[LP and Bednar, 2015]

Sparse representations in computer vision

[LP, 2021]

Sparse representations in neuromorphic engineering

Ultimately, we get a list of events for each pixel that can be merged to represent the entire image. This list of events includes pixel addresses, times of occurrence, and polarities. Note that since events are generated over time, they are naturally sorted by their time of occurrence. These events are then transmitted in real time to the output bus, often via a USB3 connection. It’s interesting to draw a parallel between this process and the optic nerve that connects our retina to the brain. In fact, the output of the retina consists of a million ganglion cells that emit action potentials, which are the only source of information transmitted by the optic nerve.

https://www.researchgate.net/profile/Guido-Croon/publication/313221316/figure/fig2/AS:668997448134663@1536512829861/Picture-of-the-event-based-camera-employed-in-this-work-the-DVS_W640.jpg

Sparse representations in neuroscience

[Brunel, 2001]

Sparse representations in neuroscience

[Mainen & Sejnowski, 1995]

Sparse representations in neuroscience

Sparse representations in neuroscience

Sparse representations in neuroscience

Sparse representations?

Sparse representations in a nutshell

…let’s delve into a computational theory of sparse coding review_bib = s.content_bib(“LP”, “2015”, ‘“Sparse models” in Biologically Inspired Computer Vision’)

Sparse representations in a nutshell

Sparse representations in a nutshell

Sparse representations in a nutshell

Generative model of image synthesis:

$I[x, y] = $ $\sum_{i=1}^{K} a[i] \cdot \phi[i, x, y]$ $ + \varepsilon[x, y]$ Where $\phi$ is a dictionary of $K$ atoms, $a$ is a sparse vector of coefficients, and $\varepsilon$ is a noise term.

Sparse representations in a nutshell

Sparse representations in a nutshell

Given an observation $I$,

$$ \begin{aligned} \mathcal{L}(a) & = - \log Pr( a | I ) \\ \end{aligned} $$

Sparse representations in a nutshell

Given an observation $I$,

$$ \begin{aligned} \mathcal{L}(a) & = - \log Pr( a | I ) \\ & = - \log Pr( I | a ) - \log Pr(a) \\ \end{aligned} $$

Sparse representations in a nutshell

Given an observation $I$,

Sparse representations in a nutshell

The problem is formalized as an optimization problem $a^\ast = \arg \min_a \mathcal{L}(a)$ with:

$$ \mathcal{L} = \frac{1}{2} \sum_{x, y} ( I[x, y] - \sum_{i=1}^{K} a[i] \cdot \phi[i, x, y])^2 + \lambda \cdot \sum_i ( a[i] \neq 0) $$

Sparse representations in a nutshell

The problem is formalized as an optimization problem $a^\ast = \arg \min_a \mathcal{L}(a)$ with:

$$ \mathcal{L}(a) = \frac{1}{2} \sum_{x, y} ( I[x, y] - \sum_{i=1}^{K} a[i] \cdot \phi[i, x, y])^2 + \lambda \cdot \sum_{i=1}^{K} | a[i] | $$

Sparse representations in a nutshell

[Rentzeperis et al (2023)]

Sparse representations in a nutshell

Neural implementation = gradient descent

LASSO = least absolute shrinkage and selection operator

Orthogonal Matching Pursuit (OMP): OMP is an iterative algorithm used for sparse signal recovery. It starts with an initial sparse solution and iteratively selects the most correlated dictionary atoms with the residual signal. OMP aims to minimize the L2 norm of the residual while maintaining sparsity. It has a greedy nature and can provide a near-optimal sparse solution.

Basis Pursuit (BP): Basis Pursuit is an optimization problem that seeks the sparsest solution to an underdetermined linear system of equations. It involves minimizing the L1 norm of the coefficient vector subject to a linear constraint. BP can be solved using linear programming techniques or convex optimization algorithms.

Iterative Soft Thresholding Algorithm (ISTA): ISTA is an iterative optimization algorithm commonly used in sparse coding. It alternates between a gradient descent step and a soft thresholding step. The gradient descent step minimizes the data fidelity term, and the soft thresholding step enforces sparsity by setting small coefficients to zero. ISTA converges to a sparse solution and can be used for dictionary learning.

FISTA (Fast Iterative Shrinkage-Thresholding Algorithm): FISTA is an accelerated version of ISTA that improves convergence speed. It incorporates momentum into the optimization process and achieves faster convergence rates.

ADMM (Alternating Direction Method of Multipliers): ADMM is an optimization technique that decomposes the original problem into smaller subproblems and solves them iteratively. It is often used for convex optimization problems with L1 regularization. ADMM has been applied to solve sparse coding problems efficiently.

Matching pursuit algorithm

Init : Residual $R = I$, sparse vector $a$ such that $\forall i$, $a[i] = 0$
while $\frac{1}{2} \sum_{x, y} R[x, y]^2 > \vartheta $, do :

Matching pursuit algorithm

Init : $R = I$, $\forall i$, $a[i] = 0$
while $\frac{1}{2} \sum_{x, y} R[x, y]^2 > \vartheta $, do :
- compute $c[i] = \sum_{x, y} (R[x, y] - a[i] \cdot \phi[i, x, y])^2$
- Match: $i^\ast = \arg \min_i c[i]$
  greedy, one by one

Matching pursuit algorithm

Init : $R = I$, $\forall i$, $a[i] = 0$
while $\frac{1}{2} \sum_{x, y} R[x, y]^2 > \vartheta $, do :
- Match : $i^\ast = \arg \max_i \sum_{x, y} R[x, y] \cdot \phi[i, x, y]$
  use of correlation instead of energy assign th first value of the sparse vector to the winning one

Matching pursuit algorithm

Init : $R = I$, $\forall i$, $a[i] = 0$
while $\frac{1}{2} \sum_{x, y} R[x, y]^2 > \vartheta $, do :
- Match : $i^\ast = \arg \max_i \sum_{x, y} ( I[x, y] \cdot \phi[i, x, y])$
- Assign : $a[i^\ast] = \frac{\sum_{x, y} R[x, y] \cdot \phi[i^\ast, x, y]}{\sum_{x, y} \phi[i^\ast, x, y] \cdot \phi[i^\ast, x, y]}$

Matching pursuit algorithm

Init : $R = I$, $\forall i$, $a[i] = 0$, and normalize $\sum_{x, y} \phi[i, x, y]^2 = 1$
while $\frac{1}{2} \sum_{x, y} R[x, y]^2 > \vartheta $, do :
- Match : $i^\ast = \arg \max_i \sum_{x, y} R[x, y] \cdot \phi[i, x, y]$
- Assign : $a[i^\ast] = \sum_{x, y} R[x, y] \cdot \phi[i^\ast, x, y]$

Matching pursuit algorithm

Init : $R = I$, $\forall i$, $a[i] = 0$, $\sum_{x, y} \phi[i, x, y]^2 = 1$
while $\frac{1}{2} \sum_{x, y} R[x, y]^2 > \vartheta $, do :
- Match : $i^\ast = \arg \max_i \sum_{x, y} R[x, y] \cdot \phi[i, x, y]$
- Assign : $a[i^\ast] = \sum_{x, y} R[x, y] \cdot \phi[i^\ast, x, y]$
- Pursuit : $R[x, y] \leftarrow R[x, y] - a[i^\ast] \cdot \phi[i^\ast, x, y]$

Matching pursuit algorithm

Init : $R = I$, $\forall i$, $a[i] = 0$, $\sum_{x, y} \phi[i, x, y]^2 = 1$
compute $c[i] = \sum_{x, y} R[x, y] \cdot \phi[i, x, y]$
compute $X[i, j] = \sum_{x, y} \phi[i, x, y] \cdot \phi[j, x, y]$
while $\frac{1}{2} \sum_{x, y} R[x, y]^2 > \vartheta $, do :
- Match : $i^\ast = \arg \max_i c[i]$
- Assign : $a[i^\ast] = c[i^\ast]$
- Pursuit : $c[i] \leftarrow c[i] - a[i^\ast] \cdot X[i, i^\ast] $

[LP (2004)]

use of correlation assign th first value of the sparse vector to the winning one

Matching pursuit algorithm

Matching pursuit algorithm

Hebbian learning (once the sparse code is known):

$$ \phi_{i}[x, y] \leftarrow \phi_{i}[x, y] + \eta \cdot a[i] \cdot (I[x, y] - \sum_{i=1}^{K} a[i] \cdot \phi_{i}[x, y] ) $$ [LP (2015)]

Unsupervised Learning of the dictionary

Hebbian learning

Matching pursuit algorithm

Sparse representations in a nutshell

Convolutional Sparse Coding

Convolutional Neural Nets (CNN)

Convolutional Neural Nets (CNN)

[Jérémie & LP, 2023]

sota…

Convolution: Mathematics

One-dimensional discrete convolution (eg in time) with a kernel $g$ of radius $K$: $$ (f \ast g)[n]=\sum_{m=-K}^{K} f[n-m] \cdot g[m] $$

Convolution: Mathematics

Convolution of an image (two-dimensional) with a kernel $g$ of radius $K\times K$:

$$ (f \ast g)[x, y] = \sum_{i=-K}^{K} \sum_{j=-K}^{K} f[x-i, y-j] \cdot g[i, j] $$

Convolution: Mathematics

Cross-correlation of an image (two-dimensional) with a kernel $g$ of radius $K\times K$:

$$ (f \ast \tilde{g})[x, y] = \sum_{i=-K}^{K} \sum_{j=-K}^{K} f[x+i, y+j] \cdot g[i, j] $$

Convolution: Mathematics

[Amidi & Amidi]

Convolution: Mathematics

Correlation of an image defined on several channels (note the order of the indices):

$$ (f \ast \tilde{g})[x, y] = \sum_{c=1}^{C} \sum_{c,i,j} f[c, x+i, y+j] \cdot g[c, i, j] $$

Convolution: Mathematics

Correlation of a multi-channel image for multiple output channels (note the order of the indices):

$$ (f \ast \tilde{g})[k, x, y] = \sum_{c,i,j} f[c, x+i, y+j] \cdot g[k, c, i, j] $$

now we get to the full CNN

CNN: the HMAX model

[Serre and Poggio, 2006]

CNN: challenges

Convolutional Sparse Coding

Convolutional Sparse Coding

Convolutional Sparse Coding

[LP, 2015]

Code @ SparseEdges

Convolutional Sparse Coding

[Ladret et al, 2024]

Convolutional Sparse Coding

Convolutional Sparse Coding

novel challenges for CNNs

1/ backpropagation is not bioplausible

CNN: Predictive processing

CNN: Predictive processing

CNN: Predictive processing

CNN: Predictive processing

From: https://pressbooks.umn.edu/sensationandperception/chapter/columns-and-hypercolumns-in-v1/

CNN: Predictive processing

CNN: Topography

[Bosking et al, 1997]

CNN: Topography

[Boutin et al, 2022]

---

Sparse representations

Laurent Perrinet

NeuroSchool PhD Program in Neuroscience

[2025-03-11]

Code / Contact me @ laurent.perrinet@univ-amu.fr

2025-02-11-neuromath

Tue, 11 Feb 2025 00:00:00 +0000

[2025-02-11] When Cortical Neurons Talk Sideways: Beyond Feedforward Visual Processing

Laurent Perrinet - https://laurentperrinet.github.io Séminaire Neuromathématiques, Collège de France

Hi, thanks for the introduction! I am Laurent Perrinet, a researcher in computational neuroscience and currently a research director at CNRS at the Institute of Neuroscience of la Timone in Marseille. Thank you for inviting me to participate in this “NeuroMathematics” seminar at the intersection of mathematics and neuroscience.

As an engineer by training, I could have pursued a career in aeronautics rather than becoming a neuroscientist. It is thanks to my mathematics professor Manuel Samuelides that I discovered the beauty of neural networks at the end of my engineering studies. This developped a curiosity, and thanks to him, I was also able to study in a mastere of cognitive sciences (now called CogMaster) in 1998. This is where I particularly want to acknowledge Jean Petitot - for his course I discovered how natural image statistics could link to principles in the central nervous system. This was a vivid revelation, and I’m grateful for his guidance in my academic path. Today’s seminar represents a return to these roots, as I’ll present my research progress since my mastere thesis on this very topic.

Today, I will address our current knowledge about horizontal connectivity rules in V1. Why is this important? As a matter of fact, one main function of sensory systems, such as the pivotal role of the primary visual cortex for vision, is to bind together the different visual features to help ultimately build a global perception.

Paysage catalan (Le Chasseur), Joan Miró (1924)

Trames (Étienne Rey)

Contour detection and the Association Field

[Field et al, 1993]

Contour detection and the Association Field

[Field et al, 1993]

Contour detection and the Association Field

[Field et al, 1993]

Natural Images : Edges are on a common circle

[Sigman et al, 2001]

A significant contribution to understanding the association field came from studying edge co-occurrences in natural images by Sigman et al. (2001). They quantified the probability density function of edge co-occurrences based on their relative positions and orientations. The figure demonstrates this by showing the spatial distribution patterns for edges relative to a reference edge at different orientations. For iso-oriented edges (a), the co-occurrence pattern shows clear structure. As the relative orientation increases through 22.5° (b), 45° (c), 67.5° (d), to 90° (e), distinct spatial patterns emerge.

A key finding was that for any given relative orientation between edges, the angle of maximal interaction occurs at the bisector between the orientations. This suggests that co-occurring edges tend to lie on a common circle - a property known as cocircularity. Panel (f) illustrates this geometrical principle: given two edges at angles w (red, 20°) and c (blue, 40°), the cocircularity solutions (green lines at 30° and 120°) represent the possible orientations of connecting circular arcs. This mathematical relationship provides insights into how the visual system might leverage statistical regularities in natural scenes for contour integration. We will go back into the details of this a bit further in the talk.

This association field concept provided a compelling framework for understanding how the visual system may implement contour integration through neural connectivity patterns. but before going there we should go back to the basic anatomy of the visual cortex.

Contour detection and the Association Field

[Field et al, 1993]

Dynamics of vision

Human Visual system (Grimaldi et al 2022)

Anatomy of the Human Visual system

Human Visual system (Grimaldi et al 2022)

Thalamic, short- & long-range lateral, inter-areal

A key feature of primary visual cortex is its layered organization, which is shared across cortical areas. The main thalamic input arrives in layer 4, which connects to a dense network of vertical connections across layers. These columns can then communicate via horizontal connections within layers. Hubel and Wiesel also proposed the ice-cube model that every point in the visual field produces a response in a 2 mm x 2 mm area of the cortex. Such an area can contain two complete groups of ocular dominance columns, 16 blobs and interblobs that may contain more than two times all of the orientations possible across 180 degrees. This region of the cortex, which Hubel and Wiesel called a hypercolumn (or, more generally, a cortical module) seems both necessary and sufficient for analyzing the image of a point in visual space. Because the cortex is a continuous cellular layer and because it is very hard to establish the boundaries of these modules physically, their existence from a functional standpoint is still the subject of debate. https://thebrain.mcgill.ca/flash/a/a_02/a_02_cl/a_02_cl_vis/a_02_cl_vis.html

Figure 9.2. Hypercolumn Diagram. Ocular dominance columns are segregated into left and right eye inputs. Orientation columns are neurons that get excited at different orientations and a cluster of these is called a pinwheel. Blobs are color selective and for every pinwheel there is a blob. (Credit: McGill: The Brain from Top to Bottom, Figure of hypercolumns, Copyleft https://copyleft.org/, https://thebrain.mcgill.ca/flash/a/a_02/a_02_cl/a_02_cl_vis/a_02_cl_vis.html. No modifications.)

Thalamic, short- & long-range lateral, inter-areal

[Markov et al 2011]

Anatomy of the Primary Visual Cortex

[Kaschube et al (2010)]

V1 is central to these pathways and shows distinctive anatomical and functional properties along with a complex topographical organization.

This figure from Kaschube et al. (2010) illustrates the organization of orientation preference maps in primary visual cortex (V1). Individual V1 neurons exhibit selective responses to oriented visual stimuli (as denoted by varying hues Colors code preferred ORs as indicated by the bars in (C)), with their spatial arrangement following highly structured patterns across the cortical surface. Panel B shows Synthetic orientation-maps of equal column spacing Λ but widely different pinwheel densities ρ. Left to right: solutions of different models: (13–16).. (C) High (blue frame) and low (orange frame) pinwheel density regions in tree shrew visual cortex. (D to F), Optically recorded orientation-maps in tree shrew (D), galago (E), and ferret (F) visual cortex. Regions shown in (C) are marked in (D). White arrows in (F) mark selected pinwheel centers. Framed regions in (C) and (F) are magnified. In many mammals including cats, monkeys and ferrets, orientation preference is organized in a quasi-periodic manner, forming what are known as orientation preference maps. These maps show remarkable consistency in their geometric properties across species, particularly in the spatial organization of pinwheel centers where orientation preferences converge.

However, this organization shows important species-specific variations. Most notably, while primates and carnivores display orderly orientation maps with smooth transitions between preferred orientations, rodents lack such maps and instead show a “salt-and-pepper” arrangement where neighboring neurons have seemingly random orientation preferences. This organizational diversity raises interesting questions about the computational advantages of these different architectures and their relationship to visual processing requirements and behavioral needs across species.

Horizontal connectivity links different hypercolumns

[Bosking et al, 1997]

This figure shows landmark results by Bosking et al. (1997) combining orientation preference maps with retrograde tracers. After injecting tracers (white arrow), they found labeled synapses (black dots) primarily connecting neurons of similar orientation preference, leading to the influential “like-to-like” connectivity hypothesis. However, later studies by Hunt, Goodhill and others revealed significant diversity in these connection patterns across cortical regions and species, suggesting more complex connectivity rules than initially proposed. This nuanced understanding has important implications for how we think about the functional organization of horizontal connections in V1.

Contour detection and the Association Field

[Field et al, 1993]

The like-to-like hypothesis

[Field et al, 2013]

The resemblance between what was shown by Bosking and the structure of the association field that we saw above is such that it is tempting to align both and state that the function of horizontal connections is to bind neurons with a selectivity to *similar orientations over long distances. This like-to-like hypothesis has been influential in understanding horizontal connectivity patterns.

However, we should be cautious about overstating these relationships. While horizontal connections show some orientation specificity, recent evidence indicates the connectivity patterns are more complex and heterogeneous than initially proposed. The functional role of this diverse connectivity remains an active area of investigation.

During the remainder of this talk**, I will try to shed light on our current knowledege on horizontal connectivities.

Supplementary: the HMAX model

[Serre and Poggio, 2007]

Supplementary: Convolutional Neural Nets (CNN)

[Chavane, LP and Rankin, 2022]

Supplementary: Orientation selectivity in V1

[Hubel & Wiesel, 1962]

Supplementary: Orientation selectivity in V1

[Hubel & Wiesel, 1962]

Supplementary: Marr’s three levels of analysis

Marr, 1982

cut in different levels: Marr (+ Poggio)
arbitrary, but useful division of labor= computational / algorithm / hardware
here:
anatomy
algorithm / model
function

First: What is the anatomy of horizontal connections?

<!--

[Marr, 1982]]

-->

Challenging the like-to-like hypothesis

Challenging the like-to-like hypothesis

Revisiting Horizontal Connectivity Rules in V1: From like-to-like towards like-to-All [Chavane, LP and Rankin, 2022]

This figure illustrates different hypothetical connectivity rules for horizontal connections in V1. The target neuron (large circle on left) has a specific orientation preference indicated by its color. Following the classical like-to-like hypothesis (shown in panel A), this neuron would preferentially connect to other neurons with matching orientation preference (similar colors) across multiple hypercolumns, as indicated by the vertical red arrows. The radial spread of connections spans approximately three hypercolumns, consistent with anatomical observations. Each hypercolumn contains a complete set of orientation preferences, represented by the different colored neurons.

This first schematic (noted A) represents one of the like-to-like connectivity rules, where horizontal connections strictly follow orientation similarity.

Challenging the like-to-like hypothesis

Revisiting Horizontal Connectivity Rules in V1: From like-to-like towards like-to-All [Chavane, LP and Rankin, 2022]

Challenging the like-to-like hypothesis

Revisiting Horizontal Connectivity Rules in V1: From like-to-like towards like-to-All [Chavane, LP and Rankin, 2022]

Challenging the like-to-like hypothesis

Revisiting Horizontal Connectivity Rules in V1: From like-to-like towards like-to-All [Chavane, LP and Rankin, 2022]

Challenging the like-to-like hypothesis

Let’s first shows some functional evidence.

This video shows voltage-sensitive dye imaging (VSDI) data from cat primary visual cortex (area 17) in response to a local oriented grating stimulus. The visualization reveals two key aspects:

The broader activation pattern shown by overall fluorescence changes (gray)
The more restricted orientation-selective response pattern (colored regions)

Two contours are overlaid: a red line marking the boundary of significant activation, and a white line delineating regions with statistically significant orientation selectivity. The orientation selectivity is encoded by color hue.

The bottom plots quantify the spatiotemporal dynamics by showing:

Left: The total activated cortical area over time
Right: The extent of orientation-selective regions over time

Together, these measurements demonstrate how orientation-selective signals propagate laterally beyond the classical feedforward input zone through horizontal connections, while maintaining some degree of feature selectivity.

Challenging the like-to-like hypothesis

Revisiting Horizontal Connectivity Rules in V1: From like-to-like towards like-to-All [Chavane, LP and Rankin, 2022]

This figure shows spatial and temporal dynamics of orientation selectivity in cat V1 analyzed from voltage-sensitive dye imaging data. Panel A displays a cortical orientation map averaged over the final 145ms of the response, where hue indicates preferred orientation and brightness shows orientation tuning strength. The dotted red line delineates the expected retinotopic boundary of feedforward input based on Albus (2004).

The inset quantitatively compares the spatial extent of:

Total cortical activation (grey contour)
Orientation-selective activation (black contour)
Theoretical feedforward input boundary (red contour)

This data demonstrates that orientation-selective responses propagate laterally beyond the classical feedforward input zone through horizontal connections, while maintaining some degree of feature selectivity. The systematic comparison between total activation and selective activation provides direct evidence for how horizontal connectivity shapes the spatiotemporal dynamics of orientation processing in V1.

Challenging the like-to-like hypothesis

Revisiting Horizontal Connectivity Rules in V1: From like-to-like towards like-to-All [Chavane, LP and Rankin, 2022]

Challenging the like-to-like hypothesis

Revisiting Horizontal Connectivity Rules in V1: From like-to-like towards like-to-All [Chavane, LP and Rankin, 2022]

Panel C displays intracellular recordings of subthreshold responses visualized as a visuotopic orientation polar map. The color hue represents preferred orientation while brightness indicates the strength of orientation tuning in the membrane potential. White contours outline regions showing statistically significant responses based on both amplitude and orientation selectivity criteria. The middle plots show averaged subthreshold responses to four different oriented stimuli (color-coded) at specific recording locations (marked by circle, triangle and square symbols), with scale bars indicating 50 ms and 1 mV. On the right, normalized orientation tuning curves are shown, computed by integrating responses within a fixed temporal window (shaded region in middle panel). The black circle marks the spontaneous activity level for the depolarizing integral measurement.

These shows a direct functional evidence for a diversity of tuning profile in th horizontal connectivity.

Challenging the like-to-like hypothesis

[Voges and LP, 2012]

Challenging the like-to-like hypothesis

Revisiting Horizontal Connectivity Rules in V1: From like-to-like towards like-to-All [Chavane, LP and Rankin, 2022]

Figure 4 illustrates a neural field model that bridges anatomical structure with functional observations in V1, as developed by Rankin and Chavane (2017).

Panel A depicts radial connectivity profiles with Gaussian-decaying inhibition and distance-dependent excitation that peaks periodically at multiples of distance L. The Ring Width (RW) parameter controls the spread of these excitatory peaks.

Panel B shows how local orientation preference maps influence lateral connectivity patterns under different orientation bias (BR) values in the recurrent connections.

Panel C quantifies the orientation tuning that emerges from these connectivity patterns. While orientations are uniformly represented globally, the local excitatory component shows strong bias around -60°. As BR increases above 0.5, the lateral connection orientation bias strengthens, reaching values around k=1 (consistent with Buzás et al. 2006).

Challenging the like-to-like hypothesis

Revisiting Horizontal Connectivity Rules in V1: From like-to-like towards like-to-All [Chavane, LP and Rankin, 2022]

Challenging the like-to-like hypothesis

Revisiting Horizontal Connectivity Rules in V1: From like-to-like towards like-to-All [Chavane, LP and Rankin, 2022]

Panel F maps the normalized selective area (relative to the feedforward footprint) across Ring Width (RW) and orientation bias (BR) parameters. White contours delineate anatomically plausible ranges where k values fall between 0.7-1.2, consistent with experimental measurements. The green region indicates parameter combinations that additionally satisfy constraints on both orientation preference and the observed radial decay of selectivity.

The neural field model effectively connects anatomical connectivity patterns with functional observations of orientation selectivity propagation in V1. The resulting connectivity structure exhibits similarities with “association field” patterns, suggesting potential optimization for encoding natural image statistics. This framework provides a quantitative basis for investigating computational principles underlying horizontal connectivity in visual cortex.

Challenging the like-to-like hypothesis

Revisiting Horizontal Connectivity Rules in V1: From like-to-like towards like-to-All [Chavane, LP and Rankin, 2022]

Modelling the Association field

[Field et al, 2013]

Edge co-occurences in natural images

Revisiting Horizontal Connectivity Rules in V1: From like-to-like towards like-to-All [Chavane, LP and Rankin, 2022]

Panel A illustrates the groundbreaking approach developed by Geisler et al. (2001) for analyzing edge statistics in natural images. The method involves:

Detecting oriented edge elements in natural images (shown as red segments)
For each edge pair, measuring:
- Their relative orientation difference (𝜃)
- The relative position angle (𝜙)
- Center-to-center distance (d)
- Azimuth difference (φ)
- Co-circularity parameter ψ = φ - θ/2

This quantitative analysis reveals two key distributions:

A strong bias for parallel edge arrangements, evident in the orientation difference histogram
A marked preference for co-circular alignments, shown in the relative position histogram

These statistics vary significantly across image databases. For example, images containing animals exhibit enhanced co-circularity compared to general natural scenes. This suggests that rather than implementing a single fixed association field, the visual system may need to handle diverse statistical regularities present in natural inputs.

The next section will examine how these statistical regularities inform computational models of the association field.

Edge co-occurences in natural images

[Geisler, 2001]

Edge co-occurences in natural images

Edge co-occurrences can account for rapid categorization of natural versus animal images [LP and Bednar, 2015]

The probability distribution function p(ψ,θ) represents the distribution of the different geometrical arrangements of edges’ angles, which we call a “chevron map”. We show here the histogram for non-animal natural images, illustrating the preference for co-linear edge configurations. For each chevron configuration, deeper and deeper red circles indicate configurations that are more and more likely with respect to a uniform prior, with an average maximum of about 3 times more likely, and deeper and deeper blue circles indicate configurations less likely than a flat prior (with a minimum of about 0.8 times as likely). Conveniently, this “chevron map” shows in one graph that non-animal natural images have on average a preference for co-linear and parallel edges, (the horizontal middle axis) and orthogonal angles (the top and bottom rows), along with a slight preference for co-circular configurations (for ψ =0 and ψ = ± π/2, just above and below the central row).

Edge co-occurences in natural images

Edge co-occurrences can account for rapid categorization of natural versus animal images [LP and Bednar, 2015]

Edge co-occurences in natural images

Edge co-occurrences can account for rapid categorization of natural versus animal images [LP and Bednar, 2015]

This figure shows classification performance across image categories using different statistical features. We used an SVM classifier with three feature sets: first-order orientation statistics (FO), the reduced 2D “chevron map” (CM), and full 4D second-order statistics (SO). The classification accuracy (F1 score) was tested for distinguishing between image categories. Results show strong performance in separating man-made from natural images, as expected. More notably, the classifier achieved ~80% accuracy in discriminating animal vs non-animal natural images, matching human performance levels reported by Serre et al. This suggests that relatively simple edge co-occurrence statistics contain sufficient information for basic image categorization tasks, without requiring higher-level semantic processing.

We also found that our model made the same errors as humans do: if an image without an animal contains more co-circular edges, it is more likely to be falsely categorized as containing an animal.

Edge co-occurences in natural images

Edge co-occurrences can account for rapid categorization of natural versus animal images [LP and Bednar, 2015]

Can we explain the diversity ?

[Bosking et al, 1997]

Indeed, this diversity is revealed in the anatomical data: V1 horizontal connectivity exhibits more complexity than suggested by the classical like-to-like hypothesis. While orientation-specific connections exist, they coexist with non-selective connections that link neurons irrespective of their tuning preferences. This diversity likely serves multiple computational functions:

Specific connections could support contour integration and feature binding
Non-selective connections may enable broad contextual modulation
Mixed connectivity patterns could help maintain network stability while preserving functional specificity

This anatomical heterogeneity aligns with V1’s role in both specialized feature detection and broader contextual processing. Understanding how these distinct connectivity patterns interact remains an active area of research in visual neuroscience.

Predictive processing

To understand the diversity in horizontal connectivity patterns, we developed a biologically plausible hierarchical model based on Convolutional Neural Networks (CNNs) backbone. The model processes natural images through multiple convolutional layers organized in a hierarchical structure:.

Natural image as input

Local receptive fields via convolution operations

Hierarchical processing through multiple layers

Predictive processing

To bridge the gap between anatomical observations and functional requirements of visual processing, We added two key ingredients in the sparse deep predictive coding (SDPC) model :

Sparse connectivity patterns:

Enforcing regularization of the activity map using L1 penalty

Activity computed via recurrent local connectivity

Similar to biological observations

Feedback from efferent layers:

Predicts activity of afferent layer

Only residual prediction error is processed

Defines long-range inter-areal connectivity

Specific influence demonstrated in Neural Computation paper

By defining a cost on minimizing the prediction error in each layer, everything stays derivable, such that we can use a classical gradient descent. These additions should allow us to better understand how feedback shapes visual processing in biological neural networks.

Predictive processing

Predictive processing

More specifically in the context of our focus today, we can look at the co-occurence

llustration of the procedure to generate interaction map. In this illustrative example we consider a V1 representation with only 4 feature maps (represented in the upper-left box). Step 1 is to extract a neighborhood (of size 3x3 in the illustration only) around the most strongly activated neuron (represented with a red square in the illustration) for a given central preferred orientation (denoted ✓ c ). Step 2 is to normalize the neural activity in the extracted neighborhood using the marginal activity (see Eq.8). Step 3 is to compute the resulting orientation and activity at every position of the neighborhood using a circular mean (see Eq. 11 and Eq. 12 respectively). To keep a concise figure we have illustrated the computation of the central edge of the interaction map only. For simplification, the illustration shows only 1 neighborhood extraction whereas the interaction maps shown in the paper are computed by averaging neighborhoods centered on the 10 most strongly activated neurons

Predictive processing

Predictive processing

Predictive processing with pooling

[Boutin et al, 2022]

## Challenging the like-to-like hypothesis

[Bosking et al, 1997]

Challenging the like-to-like hypothesis

Revisiting Horizontal Connectivity Rules in V1: From like-to-like towards like-to-All [Chavane, LP and Rankin, 2022]

To conclude, our review of horizontal connectivity in V1 reveals patterns more complex than initially theorized. The classical like-to-like hypothesis, while valuable, doesn’t fully capture the diversity of observed connectivity patterns.

Mathematical modeling has proven essential in bridging theory and biology. Our predictive processing framework shows how simple computational principles can explain the emergence of these complex connectivity patterns. The model demonstrates how feedback influences lateral interactions and reproduces key experimental observations.

However, important questions remain unanswered. We need to better understand how precise timing information is encoded in these circuits, how temporal dynamics shape processing, and whether similar principles apply across other cortical areas.

These fundamental questions will guide future experimental and theoretical work as we continue to unravel the computational principles of cortical processing.

[2025-02-11] When Cortical Neurons Talk Sideways: Beyond Feedforward Visual Processing

Laurent Perrinet - https://laurentperrinet.github.io Séminaire Neuromathématiques, Collège de France

Dynamics of vision

[Marr, 1982]

Dynamics of vision

Visual latencies (see review).

Dynamics of vision

Sensorimotor delays (Perrinet & Friston 2014)

Dynamics of vision

Sensorimotor delays (Perrinet & Friston, 2014)

Dynamics of vision

Diagonal markov model (Khoei et al, 2017).

Dynamics of vision

Flash-lag effect: MBP (Khoei et al, 2017)

Dynamics of vision

[Marr, 1982]

Dynamics of vision: Neural modeling

Spiking Neural Networks: Spiking motifs

[Grimaldi et al, 2023, Precise Spiking Motifs]

Spiking Neural Networks: Spiking motifs

Review on Precise Spiking Motifs.

Spiking Neural Networks: Spiking motifs

Review on Precise Spiking Motifs.

Spiking Neural Networks: HD-SNN

2 MINUTE

We then used a neural network with a classical architecture, which we enhanced by using an impulse representation that takes into account different possible synaptic delays. In this figure, we have represented the input in the left grid, which represents the occurrence of spikes of positive or negative polarity. Then we have represented different processing channels denoted by the colors green and orange, which are applied to this input to produce membrane activity. As illustrated above, this activity will produce output pulses, notably in synaptic connection nuclei, with heterogeneous delays corresponding to the detection of precise spatio-temporal patterns.

Spiking Neural Networks: HD-SNN

2 MINUTE

One advantage of this network is that it is differentiable, enabling us to apply classical machine learning methods, notably supervised learning. We then see the emergence of different convolution kernels, and here I represent a subset of its kernels for different directions, as denoted by the red arrows on the left of the graph. It shows the kernels obtained on the spatial representation according to the different columns, and each row represents the different delays from a delay of one on the right to a delay of 12 time steps on the left. Detectors that follow the motion emerge. For example, for the top line from top to bottom. These kernels integrate both positive neurons in red and negative polarity inputs in blue. Such spatio-temporal filtering is observed in neurobiology, but to my knowledge had never been observed in a model of spiking neurons trained under natural conditions.

Spiking Neural Networks: HD-SNN

Spiking Neural Networks: HD-SNN

Spiking Neural Networks: HD-SNN

How and why foveated retinotopy provides efficient vision

Wed, 08 Jan 2025 00:00:00 +0000

When: Wednesday 9th of January, 2025 at 12 noon.
Where: CRN seminar room, Montreal General Hospital, Livingston Hall, L7-140, with hybrid option.
Read the corresponding paper
Foveated Retinotopy Improves Classification and Localization in CNNs
Jean-Nicolas Jérémie, Emmanuel Daucé, Laurent U Perrinet

PDF Cite DOI URL arXiv

Integrating the What and Where Visual Pathways to Improve CNN Categorisation

Wed, 01 Jan 2025 00:00:00 +0000

🧠 TODAY at #CCN2025 ! Poster A145, 1:30-4:30pm at de Brug & E‑Hall. We’ve developed a bio-inspired “What-Where” CNN that mimics primate visual pathways - achieving better classification with less computation. Come chat! 🎯

What: Poster A145
When: Tuesday, August 12, 1:30 – 4:30 pm,
Where: CCN 2025 conference venue, de Brug & E‑Hall Presented by main author Jean-Nicolas JÉRÉMIE and in cosupervision with Emmanuel Daucé https://laurentperrinet.github.io/publication/jeremie-25-ccn/ Our research introduces a novel “What-Where” approach to CNN categorization, inspired by the dual pathways of the primate visual system:

The ventral “What” pathway for object recognition
The dorsal “Where” pathway for spatial localization Key innovations:

✅ Bio-inspired selective attention mechanism

✅ Improved classification performance with reduced computational cost

✅ Smart visual sensor that samples only relevant image regions

✅ Likelihood mapping for targeted processing The results?

Better accuracy while using fewer resources - proving that nature’s designs can still teach us valuable lessons about efficient AI. Come find us this afternoon for great discussions! #CCN2025 #ComputationalNeuroscience #AI #MachineLearning #BioinspiredAI #ComputerVision #Research

Links:

Artificial neural networks and machine learning applied to the understanding of biological vision

Mon, 13 May 2024 14:00:00 +0000

PhD thesis 'Focus of attention: a sensory-motor task for energy reduction in spiking neural networks'

Fri, 03 May 2024 09:00:00 +0000

Dear colleagues,

Applications are welcome for a fully funded PhD position Focus of attention: a sensory-motor task for energy reduction in spiking neural networks. The position will be located at the EDGE Team @ LEAT Laboratory within Université Côte d’Azur and/or at the INT in Marseille, France.

Context

This project takes place in the context of the EMERGENCES project (ANR PEPR IA 2023-2027) which aims to advance the state of the art on machine learning models using inspiration from biology. Indeed, inspiration from brain features promises to show the emergence of unrivalled efficient processing. Among the most promising features studied in the literature of bio-inspired AI are temporal data encoding using spikes, multimodal association, local learning or attention-based processing.

This PhD subject focuses on the association between attention and spiking neural networks for defining new efficient AI models for embedded systems such as drones, robots and more generally autonomous systems.

The thesis will take place between the LEAT research lab in Sophia-Antipolis and the INT institute in Marseille which both develop complementary approaches on bio-inspired AI from neuroscience observation to embedded systems design.

Subject

The volume as well as the diversity of visual information that reaches our eyes at every moment are huge and cannot be fully integrated by the visual system. In other words, the biological system is confronted to the same challenge as the one encountered by artificial systems (especially at the edge) when dealing with the huge amounts of information coming continuously from the real world. Interestingly, the brain has found an original approach to deal with this issue by focusing on a sub-part of the visual information at a time. Indeed, the study of the visual cortex in neuroscience has made it possible to highlight subregions that treat each or all of the multiple properties of information coming from the visual pathways: shapes, colors, movements, etc [1], thus revealing the interaction of attentional processes and the concept of “saliency” used in cognitive science.

Creating a fully autonomous system remains a significant challenge, especially when operating in the dynamic real world. In recent times, machine learning has assumed a prominent role in machine vision, particularly through the implementation of deep learning algorithms. These algorithms have yielded impressive outcomes in tasks such as object detection, recognition, and tracking. However, these systems come with a high computational cost, as they must process entire camera images to generate these results. Additionally, they struggle to dynamically adapt to changes in their environment.

Our focus lies on two integrated bio-inspired approaches that leverage attentional mechanisms. The first approach, known as bottom-up, draws inspiration from the work of the Gestalt theory, the Feature Integration Theory (Triesman, Gelad) [3], and the model of visual attention from Itti & Koch [1]. This approach relies on the saliency of low-level features in the visual field, processed in parallel, including movement, color, and edges. It employs emergent mechanisms to integrate features guided by their saliency in order to detect the consistency of objects, encompassing their form, position, and speed. As shown by the Gestalt theory, only the more salient data are needed in this mechanism. Thus, we can dramatically reduce the amount of needed data by extracting only the more salient regions of interest during bottom-up phase.

The second approach, known as top down, considers that the visual attention is guided by higher level cognitive stages. For instance, in the Guided Search theory [4], Wolfe emphasizes the role of prior knowledges, expectations, and intentions. In this work, Wolfe proposes a guided search mechanism that relies on a “Priority map that represents the system’s best guess as to where to deploy attention next.”. This Priority map is built on multiple sources of information such as the visual system as well as higher-level information such as intention, search history and the actual visual semantics. In this way, higher-level information is used to guide the filtering of the botom-up path, so that only the information required for a given task is selected and processed. Similar systems are proposed by Schöner [5] in which saliency maps, working memories and “priority map”, guided visual search mechanisms are implemented through the Neural Field Theory (NFT). Here, Dynamic Neural Fields are used to implement the saliency of feature maps, as well as scene spatial selection mechanism, working memory, etc.

In a previous work from the LEAT [6], we have proposed a brain inspired attentional process implementing bottom-up and top-down paths based on a dynamic neural fields properties embodied in a sensory-motor loop. In a complementary work, the INT group has developed a dual pathway model of the visual system in which saliency emerges as a property of the perceptual system to perform saccades, that is, rapid shifts of the fixation point [7]. This uses a recognition model which takes as an input a retinotopically transformed input and shows the emergence of saliency maps [8] In the dual-pathway model, the exploration of a visual scene is based on both the saliency of the color feature (bottom-up) and the class of the last selected object recognized by a convolutional neural network (top-down). Both paths are integrated by a dynamic neural field to select the next visual information to be explored or conserved by setting motor orders accordingly.

The main goal of the thesis is to propose a new vision of the integration of attention into machine learning models. The proposed model will draw on the dynamics at play in a sensory-motor approach to perception and will thus reconsider the classical perception tasks in order to better fit with the continuous flow of information coming from the environment.

Work plan

The PhD will be co-supervised between INT in Marseille and LEAT in Nice. According to the preferences of the candidate, a main laboratory of affiliation will be selected. Weekly meetings will be organized remotely and visiting weeks will be planned to work in-person in the other lab along the year.

Year 1

Study the state of the art in both neuroscience and machine learning on the use of attentional properties to make AI models more effective in environmental perception tasks.
Write a synthesis report on this study.
Develop a first neural model integrating attention-based selection in a specific perception task such as visual search.
Define the specific metrics (KPI) dedicated to the evaluation of the performance and efficiency of such a bio-inspired AI model.
Submit a first publication on this preliminary study in an international conference.

Year 2

Analyze of the performances of the preliminary attention-based model
Develop the approach in order to integrate step by step the features related to dual pathway perception, attention, foveation, DNF and make the model compatible with convolutional neural networks
Submit a second publication in a international journal

Year 3

Study the adaptation of the model to spiking neural networks
Evaluation and comparison of the different approaches
Submit publications on the final results of the thesis
Write the thesis report and prepare the defense

Required skills

Master thesis in one of the following domains: neuromorphic systems, spiking neural networks, neurocognition, machine learning.
Background and experience in machine-learning, artificial neural networks, and/or neurosciences.
Strong motivation, team working, fluent in English (spoken and written).
Programming skills in python, keras, pytorch or equivalent

Start: year 2024

Duration: 3 years

Location: Sophia-Antipolis and/or Marseille

Contacts

Benoît Miramond is Full Professor in Electrical Engineering at LEAT laboratory from University Côte d’Azur (UCA). He holds the chair on bio-inspired AI at 3IA Cote d’Azur Institute and leads the eBRAIN research group which develops a interdisciplinary research activity on embedded Bio-inspiRed AI and Neuromorphic architectures, especially based on SNNs. LEAT is a mixt research unit (UMR 72 48) from UCA and CNRS.
Laurent Perrinet is a director of research at Institut des Neurosciences de la Timone (CNRS - Aix-Marseille Université). He is studying the link between brain microstructures and their macroscopic function by implementing realistic models of the primary visual cortex using spiking neural networks.
Laurent Rodriguez is associate professor at LEAT laboratory in the eBRAIN group. He is interested in dynamic neural networks and develop neural models from biological inspiration.
More details on the “Emergences” grant:
Laurent U Perrinet (2023). Emergences (2023 / 2027).

Application

Apply by sending an email directly to the supervisors (Benoit.miramond@univ-cotedazur.fr Laurent.perrinet@univ-amu.fr Laurent.rodriguez@univ-cotedazur.fr). The application will include:

• Letter of recommendation of the master supervisor.

• Curriculum vitæ.

• Motivation Letter.

References

[1] L. Itti et C. Koch, « Computational modelling of visual attention ». Nat Rev Neurosci, vol. 2, 3, 3, mars 2001, doi: 10.1038/35058500.
[2] Gerstner, W., Kistler, W. M., Naud, R., & Paninski, L. (2014). « Neuronal dynamics: From single neurons to networks and models of cognition ». Cambridge University Press

[3] A. M. Treisman et G. Gelade, « A feature-integration theory of attention ». Cognitive Psychology, vol. 12, 1, p. 97‑136, janv. 1980, doi:10.1016/0010-0285(80)90005-5.
[4] Wolfe, J.M. «Guided Search 6.0: An updated model of visual search ». Psychon Bull Rev 28, 1060–1092 (2021). https://doi.org/10.3758/s13423-020-01859-9
[5] Grieben, R., & Schöner, G.. « A neural dynamic process model of combined bottom-up and top-down guidance in triple conjunction visual search». In T. Fitch, Lamm, C., Leder, H., & Teßmar-Raible, K. (Eds.), Proceedings of the 43rd Annual Conference of the Cognitive Science Society
[6] M. Rasamuel, Lyes Khacef, Laurent Rodriguez, et Benoit Miramond, « Specialized visual sensor coupled to a dynamic neural field for embedded attentional process ». IEEE Conference Publication | IEEE Xplore. https://ieeexplore.ieee.org/abstract/document/8705979
[7] Emmanuel Daucé, Pierre Albigès, Laurent U Perrinet (2020). « A dual foveal-peripheral visual processing model implements efficient saccade selection ». Journal of Vision. doi:https://doi.org/10.1167/jov.20.8.22
[8] Jean-Nicolas Jérémie, Emmanuel Daucé, Laurent U Perrinet (2020). « Retinotopic Mapping Enhances the Robustness of Convolutional Neural Networks ». arXiv https://arxiv.org/abs/2402.15480

2024-04-17-phd-program-sparse-representations

Wed, 10 Apr 2024 00:00:00 +0000

Sparse representations

Laurent Perrinet

NeuroSchool PhD Program in Neuroscience

[2024-04-17]

Code / Contact me @ laurent.perrinet@univ-amu.fr

Sparse representations?

Sparse representations in computer vision

Sparse representations in computer vision

[LP and Bednar, 2015]

Sparse representations in computer vision

[LP, 2021]

Sparse representations in neuromorphic engineering

https://www.researchgate.net/profile/Guido-Croon/publication/313221316/figure/fig2/AS:668997448134663@1536512829861/Picture-of-the-event-based-camera-employed-in-this-work-the-DVS_W640.jpg

Sparse representations in neuromorphic engineering

The HD-SNN neural network.

2 MINUTE

kernels learned for motion detection

can we force a sparse connectivity (beware that’s diferent from sparse activity)

Sparse representations in neuromorphic engineering

The HD-SNN neural network.

Sparse representations in neuroscience

[Brunel, 2001]

Sparse representations in neuroscience

[Mainen & Sejnowski, 1995]

Sparse representations in neuroscience

Sparse representations in neuroscience

Sparse representations in neuroscience

Sparse representations?

Sparse representations in a nutshell

…let’s delve into a computational theory of sparse coding review_bib = s.content_bib(“LP”, “2015”, ‘“Sparse models” in Biologically Inspired Computer Vision’)

Sparse representations in a nutshell

Sparse representations in a nutshell

Sparse representations in a nutshell

Generative model of image synthesis:

$I[x, y] = $ $\sum_{i=1}^{K} a[i] \cdot \phi[i, x, y]$ $ + \varepsilon[x, y]$ Where $\phi$ is a dictionary of $K$ atoms, $a$ is a sparse vector of coefficients, and $\varepsilon$ is a noise term.

Sparse representations in a nutshell

Sparse representations in a nutshell

Given an observation $I$,

$$ \begin{aligned} \mathcal{L}(a) & = - \log Pr( a | I ) \\ \end{aligned} $$

Sparse representations in a nutshell

Given an observation $I$,

$$ \begin{aligned} \mathcal{L}(a) & = - \log Pr( a | I ) \\ & = - \log Pr( I | a ) - \log Pr(a) \\ \end{aligned} $$

Sparse representations in a nutshell

Given an observation $I$,

Sparse representations in a nutshell

The problem is formalized as an optimization problem $a^\ast = \arg \min_a \mathcal{L}(a)$ with:

$$ \mathcal{L} = \frac{1}{2} \sum_{x, y} ( I[x, y] - \sum_{i=1}^{K} a[i] \cdot \phi[i, x, y])^2 + \lambda \cdot \sum_i ( a[i] \neq 0) $$

Sparse representations in a nutshell

The problem is formalized as an optimization problem $a^\ast = \arg \min_a \mathcal{L}(a)$ with:

$$ \mathcal{L}(a) = \frac{1}{2} \sum_{x, y} ( I[x, y] - \sum_{i=1}^{K} a[i] \cdot \phi[i, x, y])^2 + \lambda \cdot \sum_{i=1}^{K} | a[i] | $$

Sparse representations in a nutshell

[Rentzeperis et al (2023)]

Sparse representations in a nutshell

Neural implementation = gradient descent

LASSO = least absolute shrinkage and selection operator

Matching pursuit algorithm

Init : Residual $R = I$, sparse vector $a$ such that $\forall i$, $a[i] = 0$
while $\frac{1}{2} \sum_{x, y} R[x, y]^2 > \vartheta $, do :

Matching pursuit algorithm

Init : $R = I$, $\forall i$, $a[i] = 0$
while $\frac{1}{2} \sum_{x, y} R[x, y]^2 > \vartheta $, do :
- compute $c[i] = \sum_{x, y} (R[x, y] - a[i] \cdot \phi[i, x, y])^2$
- Match: $i^\ast = \arg \min_i c[i]$
  greedy, one by one

Matching pursuit algorithm

Init : $R = I$, $\forall i$, $a[i] = 0$
while $\frac{1}{2} \sum_{x, y} R[x, y]^2 > \vartheta $, do :
- Match : $i^\ast = \arg \max_i \sum_{x, y} R[x, y] \cdot \phi[i, x, y]$
  use of correlation instead of energy assign th first value of the sparse vector to the winning one

Matching pursuit algorithm

Init : $R = I$, $\forall i$, $a[i] = 0$
while $\frac{1}{2} \sum_{x, y} R[x, y]^2 > \vartheta $, do :
- Match : $i^\ast = \arg \max_i \sum_{x, y} ( I[x, y] \cdot \phi[i, x, y])$
- Assign : $a[i^\ast] = \frac{\sum_{x, y} R[x, y] \cdot \phi[i^\ast, x, y]}{\sum_{x, y} \phi[i^\ast, x, y] \cdot \phi[i^\ast, x, y]}$

Matching pursuit algorithm

Init : $R = I$, $\forall i$, $a[i] = 0$, and normalize $\sum_{x, y} \phi[i, x, y]^2 = 1$
while $\frac{1}{2} \sum_{x, y} R[x, y]^2 > \vartheta $, do :
- Match : $i^\ast = \arg \max_i \sum_{x, y} R[x, y] \cdot \phi[i, x, y]$
- Assign : $a[i^\ast] = \sum_{x, y} R[x, y] \cdot \phi[i^\ast, x, y]$

Matching pursuit algorithm

Init : $R = I$, $\forall i$, $a[i] = 0$, $\sum_{x, y} \phi[i, x, y]^2 = 1$
while $\frac{1}{2} \sum_{x, y} R[x, y]^2 > \vartheta $, do :
- Match : $i^\ast = \arg \max_i \sum_{x, y} R[x, y] \cdot \phi[i, x, y]$
- Assign : $a[i^\ast] = \sum_{x, y} R[x, y] \cdot \phi[i^\ast, x, y]$
- Pursuit : $R[x, y] \leftarrow R[x, y] - a[i^\ast] \cdot \phi[i^\ast, x, y]$

Matching pursuit algorithm

Init : $R = I$, $\forall i$, $a[i] = 0$, $\sum_{x, y} \phi[i, x, y]^2 = 1$
compute $c[i] = \sum_{x, y} R[x, y] \cdot \phi[i, x, y]$
compute $X[i, j] = \sum_{x, y} \phi[i, x, y] \cdot \phi[j, x, y]$
while $\frac{1}{2} \sum_{x, y} R[x, y]^2 > \vartheta $, do :
- Match : $i^\ast = \arg \max_i c[i]$
- Assign : $a[i^\ast] = c[i^\ast]$
- Pursuit : $c[i] \leftarrow c[i] - a[i^\ast] \cdot X[i, i^\ast] $

[LP (2004)]

use of correlation assign th first value of the sparse vector to the winning one

Matching pursuit algorithm

Matching pursuit algorithm

Hebbian learning (once the sparse code is known):

$$ \phi_{i}[x, y] \leftarrow \phi_{i}[x, y] + \eta \cdot a[i] \cdot (I[x, y] - \sum_{i=1}^{K} a[i] \cdot \phi_{i}[x, y] ) $$ [LP (2015)]

Unsupervised Learning of the dictionary

Hebbian learning

Matching pursuit algorithm

Sparse representations in a nutshell

Convolutional Sparse Coding

Convolutional Neural Nets (CNN)

Convolutional Neural Nets (CNN)

[Jérémie & LP, 2023]

sota…

Convolution: Mathematics

One-dimensional discrete convolution (eg in time) with a kernel $g$ of radius $K$: $$ (f \ast g)[n]=\sum_{m=-K}^{K} f[n-m] \cdot g[m] $$

Convolution: Mathematics

Convolution of an image (two-dimensional) with a kernel $g$ of radius $K\times K$:

$$ (f \ast g)[x, y] = \sum_{i=-K}^{K} \sum_{j=-K}^{K} f[x-i, y-j] \cdot g[i, j] $$

Convolution: Mathematics

Cross-correlation of an image (two-dimensional) with a kernel $g$ of radius $K\times K$:

$$ (f \ast \tilde{g})[x, y] = \sum_{i=-K}^{K} \sum_{j=-K}^{K} f[x+i, y+j] \cdot g[i, j] $$

Convolution: Mathematics

[Amidi & Amidi]

Convolution: Mathematics

Correlation of an image defined on several channels (note the order of the indices):

$$ (f \ast \tilde{g})[x, y] = \sum_{c=1}^{C} \sum_{c,i,j} f[c, x+i, y+j] \cdot g[c, i, j] $$

Convolution: Mathematics

Correlation of a multi-channel image for multiple output channels (note the order of the indices):

$$ (f \ast \tilde{g})[k, x, y] = \sum_{c,i,j} f[c, x+i, y+j] \cdot g[k, c, i, j] $$

now we get to the full CNN

CNN: the HMAX model

[Serre and Poggio, 2006]

CNN: challenges

Convolutional Sparse Coding

Convolutional Sparse Coding

Convolutional Sparse Coding

[LP, 2015]

Code @ SparseEdges

Convolutional Sparse Coding

[Ladret et al, 2024]

Convolutional Sparse Coding

Convolutional Sparse Coding

novel challenges for CNNs

1/ backpropagation is not bioplausible

CNN: Predictive processing

CNN: Predictive processing

CNN: Predictive processing

CNN: Predictive processing

CNN: Predictive processing

CNN: Topography

[Bosking et al, 1997]

CNN: Topography

[Boutin et al, 2022]

---

Sparse representations

Laurent Perrinet

NeuroSchool PhD Program in Neuroscience

[2024-04-17]

Code / Contact me @ laurent.perrinet@univ-amu.fr

2023-11-07-snufa.md

Tue, 07 Nov 2023 00:00:00 +0000

Accurate Detection of Spiking Motifs by Learning Heterogeneous Delays of a Spiking Neural Network

Laurent Perrinet

SNUFA: Spiking Neural networks as Universal Function Approximators

^{https://laurentperrinet.github.io/talk/2023-11-07-snufa}

Core Mechanism of Spiking Motif Detection

From generating raster plots to inferring spiking motifs

A In this work, this principle was framed in a probabilistic setting such that we could provide an optimal scheme for detecting generic spiking motifs which may be superposed at random times. Starting with 10 presynaptic inputs, this model allows to generate a synthetic raster plot as the combination of four different spiking motifs. B These motifs are defined by a positive (red) or negative (blue) contribution to the spiking probability which are represented here. C Applying a Bayesian approach, we may define four formal spiking neurons which will integrate the incoming spiking information from the presynaptic neurons - this analog signal can then be thresholded to give the detection of each spiking motif (vertical) bar which was here always exact with respect to the ground truth (stars). D The beauty of this is that we can recover in the presynaptic raster plot the contribution of each spiking motif to the original raster plot.

Detecting spiking motifs using heterogeneous delays

Detecting spiking motifs using heterogeneous delays: supervised learning

Accurate Detection of Spiking Motifs by Learning Heterogeneous Delays of a Spiking Neural Network

Laurent Perrinet

SNUFA: Spiking Neural networks as Universal Function Approximators

^{https://laurentperrinet.github.io/talk/2023-11-07-snufa}

Emergences (2023 / 2027)

Thu, 05 Oct 2023 14:00:00 +0000

TL;DR: Conventional deep learning models consume too much energy. Inspired by biology, we will explore new models that are more energy efficient.

The Emergences project aims at advancing the state-of-the art on near-physics emerging models by collaboratively exploring various computation models leveraging physical devices properties. This project will focus on Event-based models, Physics-inspired models and innovative near-physics Machine Learning solutions. Emergences further intends to extend the collaborative research activities beyond the fence of the consortium by means of connecting with other projects of the PEPR IA and other research institutes.

Pilote: Marina Reyboz, CEA, Research Director
Co-Pilote: Gilles Sassatelli, CNRS, Research Director

Latest news

2026-01-29 : talk at the PEPR AI meeting
Laurent U Perrinet (2026). Neurosciences and sparsity. Séminaire au colloque du PEPR AI ``Emergences’’ 2026.
Cite Slides URL
2025-03-18: PEPR IA Days du 18 au 20 mars à CentraleSupélec.
2024-09-26: 2nd workshop meeting in Paris.
2024-05-03 : we are hiring !
Laurent U Perrinet (2024). PhD thesis 'Focus of attention: a sensory-motor task for energy reduction in spiking neural networks'.
2024-03-27 : talk at the PEPR AI meeting
Laurent U Perrinet (2024). Analyser de larges volumes de données neurobiologiques, vers une approche biomimétique.
Cite Project Slides URL
2023-10-05: Kick-off meeting!

Description of the “Emergences” project

Contemporary machine learning (ML) has incurred profound changes in the scientific, societal and economic landscapes alike. After a decade of sustained progress AI as a discipline is still making regular breakthroughs on many fronts, at the expense of an ever-increasing amount of consumption of compute resources. Modern language models feature hundreds of billion parameters and training energy consumption alone likely falls in the GWh range, with a logical forecast worsening the already prohibitive carbon footprint of AI.

Besides the flourishing initiatives aimed at defining AI-friendly digital compute stack, the next logical breakthrough on the horizon is undoubtedly the emergence of disruptive AI compute technologies having improved energy efficiency. This development will likely involve the utilization of models that differ from those traditionally used in ML and exhibit properties that resemble the behavior of physical components, thereby facilitating implementation.

The Emergences project aims at advancing the state-of-the art on near-physics emerging models by collaboratively exploring various computation models leveraging physical devices properties. Efforts will be put on 3 distinct fronts: Event-based models, Physics-inspired models (from physical systems dynamics) and innovative near-physics ML solutions (exploiting device properties). The investigations will be focused on embedded systems for Edge AI that call for increased energy efficiency for inference and learning, which could be incremental. They will apply to several application domains ranging for instance from the monitoring of the environment to health. Other important tasks such as common tools, performance metrics definition and model scalability analysis and will be conducted through as a collaborative transverse initiative.

Emergences further intends to extend the collaborative research activities beyond the fence of the consortium by means of connecting with other projects of the PEPR IA and other research institutes, some of which are listed in this proposal. Finally, because of the unavoidable societal and philosophical implications of AI as a whole, Emergences will concurrently to the research activities run a track aimed at analyzing and anticipating the impact of its upcoming contributions.

Description of the PhD project (WP1): Focus of attention: a sensory-motor task for energy reduction in unsupervised spiking neural networks

attention mechanisms based on our cognitive architecture using a dual pathway:
Emmanuel Daucé, Pierre Albigès, Laurent U Perrinet (2020). A dual foveal-peripheral visual processing model implements efficient saccade selection. Journal of Vision.
Cite DOI Pdf Code URL bioRxiv
implementation in a spiking neural network based:
Antoine Grimaldi, Laurent U Perrinet (2023). Learning heterogeneous delays in a layer of spiking neurons for fast motion detection. Biological Cybernetics.
PDF Cite DOI URL Code Supplementary Material

Carte des partenaires du projet Emergences.

L’intelligence artificielle induit des changements profonds dans les paysages scientifiques, économiques, politiques et sociétaux contemporains. Une décennie après sa « renaissance », l’apprentissage automatique continue à réaliser des avancées sur de nombreux fronts, au prix cependant d’une boulimie de ressources informatiques induisant une consommation électrique préoccupante. Les modèles de langage actuels comportent quelques centaines de milliards de paramètres et consomment pour leur entraînement seul plusieurs GWh, ce qui aujourd’hui motive la recherche d’approches (de rupture) plus sobres. En plus des diverses initiatives visant à développer des composants et systèmes numériques pensés pour l’IA et dotés d’une meilleure efficacité énergétique, des approches disruptives en IA doivent être développées pour viser des gains énergétiques encore plus importants. Cette évolution passera par l’utilisation de modèles différents de ceux utilisés traditionnellement en apprentissage et présentant des propriétés proches des comportements de composants physiques, en facilitant par là-même l’implantation. Le projet Emergences fait avancer l’état de l’art sur les modèles émergents proches de la physique en explorant de manière collaborative divers modèles de calcul en utilisant les propriétés de différents dispositifs physiques. Les efforts concentrés sur trois fronts distincts : i) les modèles événementiels bio-inspirés pour lesquels des avancées sont réalisées sur la compréhension du fonctionnement de ces modèles et leur optimisation notamment dans le cas d’implémentation sur semiconducteurs (FPGA et ASIC) mais aussi de lois d’apprentissage émergentes (Sparse Forward forward) ou encore sur la parcimonie des données ii) les modèles inspirés de la physique pour lesquelles des premières propositions concrètes de solutions permettant de réaliser des circuits émergent (création d’un testchip, modèles stochastiques, apprentissage continu bayésien, accélération matériel de couches MHA, entre autres) et enfin iii) les solutions d’apprentissage automatique innovantes proches de la physique, avec des propositions d’implémentation d’apprentissage on-chip pour des composants émergents (apprentissage forward-only sur réseaux memristifs, apprentissage multimodal et incrémental). Toutes ces investigations sont menées dans un cadre d’expérimentations basés sur des jeu de données et indicateurs de performances décidés communément, et un effort très significatif à l’endroit de la soutenabilité est réalisé avec une méthodologie d’évaluation de l’empreinte environnementale en cours de mise en place. Ainsi Emergences vise à proposer un cadre de travail pour qualifier ses propositions, face à l’état de l’art académique et industriel dans le domaine de l’IA à la périphérie (Edge AI) qui nous sert de référentiel d’analyse.

Key figures

starting date: September 1, 2023
Duration: 48 months (until August 31, 2027)
14 partners
Nb of PhD: 19
Nb of Post doc: 13
TRL: basic research
Total grant requested: 6.8 M€

This work is supported by a public grant overseen by the French National Research Agency (ANR) under the grant number ANR-23-PEIA-0002 EMERGENCES.

2023-09-27_icann.md

Wed, 27 Sep 2023 00:00:00 +0000

Accurate Detection of Spiking Motifs by Learning Heterogeneous Delays of a Spiking Neural Network

Laurent Perrinet

ICANN workshop on Recent Advances in SNNs

^{https://laurentperrinet.github.io/talk/2023-09-27-icann}

Hello, I’m Laurent Perrinet from the Institut des Neurosciences de la Timone, a joint AMU / CNRS unit, and during this talk at this ICANN workshop on Recent Advances in SNNs, I’ll be presenting a method for the Accurate Detection of Spiking Motifs by Learning Heterogeneous Delays of a Spiking Neural Network, and how it may also impact the design of SNNs. I’d like to thank Sander Bohté and Sebastian Otte for the organization of this workshop and you for listening. These slides are available from my web-site, along with a number of references. The outline of the talk is as follows: first, I’ll describe how one may perform computations using Heterogeneous Delays - and present a toy model example; then, I’ll show real scale example quantifying the performance on synthetic data ; and finally, I’ll present how this SNN is in fact differentiable and may be extended for future applications.

Core Mechanism of Spiking Motif Detection

From generating raster plots to inferring spiking motifs

Detecting spiking motifs using heterogeneous delays

Accurate Detection of Spiking Motifs by Learning Heterogeneous Delays of a Spiking Neural Network

Laurent Perrinet

ICANN workshop on Recent Advances in SNNs