Adrien F. Vincent

Line-based event preprocessing: towards low-energy neuromorphic computer vision

Amélie Gruel, Pierre Lewden, Adrien F. Vincent, Sylvain Saïghi

Abstract Neuromorphic vision has made significant progress in recent years, thanks to the natural match between spiking neural networks and event data in terms of biological inspiration, energy savings, latency and memory use for dynamic visual data processing. However, optimising its energy …

fr (code pays fourni par la source)

Line-based Event Preprocessing: Towards Low-Energy Neuromorphic Computer Vision

Amélie Gruel, Pierre Lewden, Adrien F. Vincent, Sylvain Saïghi

Neuromorphic vision made significant progress in recent years, thanks to the natural match between spiking neural networks and event data in terms of biological inspiration, energy savings, latency and memory use for dynamic visual data processing. However, optimising its energy requirements still …

Line-based Event Preprocessing: Towards Low-Energy Neuromorphic Computer Vision

Amélie Gruel, Pierre Lewden, Adrien F. Vincent, Sylvain Saïghi

Neuromorphic vision made significant progress in recent years, thanks to the natural match between spiking neural networks and event data in terms of biological inspiration, energy savings, latency and memory use for dynamic visual data processing. However, optimising its energy requirements still …

An ultra low noise, low offset chopper Operational Amplifier in 180 nm technology

Clément Dufossée, Jean-Baptiste Bégueret, Thierry Taris, Adrien F. Vincent et autres

This article describes the design of a low-complexity chopper stabilized operational amplifier (op-amp) dedicated to process the output signal of Magnetic Tunnel Junctions (MTJs). The extremely low level of DC output signal produced by MTJs requires a large voltage amplification with low …

fr (code pays fourni par la source)

A ferroelectric–memristor memory for both training and inference

Michele Martemucci, François Rummens, Yannick Malot, Tifenn Hirtzlin et autres

Developing artificial intelligence systems that are capable of learning at the edge of a network requires both energy-efficient inference and learning. However, current memory technology cannot provide the necessary combination of high endurance, low programming energy and non-destructive read processes. Here we …

fr (code pays fourni par la source)

Performance Evaluation of Neuromorphic Visual Attention Based on Proto-Objects

Amélie Gruel, Adrien F. Vincent, Chip-Hong Chang, Sylvain Saïghi

Human beings detects saliency in a visual scene using different mechanisms experimentally studied over the years, such as the detection of "proto-objects", i.e. salient regions in the visual scene where an object could plausibly be. Proto-object saliency-based models’ ability to be both …

fr, sg, gb (code pays fourni par la source)

Comparison of Hardware-friendly, Audio-to-spikes Cochlear Encoding for Neuromorphic Processing

Valentin Meunier, Amélie Gruel, Adrien F. Vincent, Sylvain Saïghi

Obtaining data from neuromorphic sensors and processing it with Spiking Neural Networks (SNNs) is a promising solution to lower energy consumption in Artificial Intelligence systems. The current rarity of natively neuromorphic datasets promotes the development of software tools to translate input sensory …

fr (code pays fourni par la source)

Insights from Towards IPOS consultation process: informing development and future directions

Lucas Becquet, Tanya Brodie Rudolph, Guillermo Ortuño Crespo, Adrien F. Vincent et autres

Insights from Towards IPOS consultation process: informing development and future directionsThe foundational value proposition for an IPOS has been significantly enriched and shaped by a methodologically diverse consultation process launched in 2022. Through a combination of one-on-one interviews, in-person and online workshops, …

fr, us, Afrique du Sud, ch, de (code pays fourni par la source)

Unified Ferroelectric/Memristive Memory for Neural Network Inference and Training

Elisa Vianello, Michele Martemucci, François Rummens, Yannick Malot et autres

fr (code pays fourni par la source)

Enseignements sur les enjeux environnementaux et les impacts du numérique dans une école d’ingénieurs

Annabelle Collin, Armande Capitaine, Adrien F. Vincent, Mathieu Coquerelle et autres

Un enseignement obligatoire de 3 h de fresque (du climat ou du numérique), de 9 h de cours magistral et de 12 h de travaux pratiques de sensibilisation aux enjeux environnementaux et aux impacts du numérique pour les élèves en première année …

fr (code pays fourni par la source)

Neuromorphic Event-based Line Detection on SpiNNaker

Amélie Gruel, Adrien F. Vincent, Jean Martinet, Sylvain Saïghi

The combined use of Spiking Neural Networks (SNNs) and neuromorphic data in recent years makes for a promising solution to the challenges currently raised in computer vision. Indeed, the natural match between SNNs and event data leads to improvements in terms of …

fr (code pays fourni par la source)

Effect of Line Resistance of Passive Memristive Crossbars on Spiking Neural Network Performance

Pierre Lewden, Adrien F. Vincent, Jean Tomas, Chip-Hong Chang et autres

Hardware spiking neural networks using memristors as synapses are promising candidates for highly integrated and low-power event-based computations. However, when using passive synaptic crossbars, a number of limitations linked to the high parallelism of the memristive synapses appear. In this paper, we …

sg, fr (code pays fourni par la source)