Petrut Bogdan

The neurobench framework for benchmarking neuromorphic computing algorithms and systems

Jason Yik, Korneel Van den Berghe, Douwe den Blanken, Younes Bouhadjar et autres

Neuromorphic computing shows promise for advancing computing efficiency and capabilities of AI applications using brain-inspired principles. However, the neuromorphic research field currently lacks standardized benchmarks, making it difficult to accurately measure technological advancements, compare performance with conventional methods, and identify promising future …

us, nl, de, se, hk, it, ch, au, gb, ca, fr, es, kr, be (code pays fourni par la source)

Interdisciplinary and Collaborative Training in Neuroscience: Insights from the Human Brain Project Education Programme

Alice Geminiani, Judith Kathrein, Alper Yegenoglu, Marcelo Armendáriz et autres

Neuroscience education is challenged by rapidly evolving technology and the development of interdisciplinary approaches for brain research. The Human Brain Project (HBP) Education Programme aimed to address the need for interdisciplinary expertise in brain research by equipping a new generation of researchers …

pt, it, at, de, us, es, gb, se, no, be (code pays fourni par la source)

NeuroBench: A Framework for Benchmarking Neuromorphic Computing Algorithms and Systems

Jason Yik, Douwe den Blanken, Younes Bouhadjar, Noah Pacik-Nelson et autres

Neuromorphic computing shows promise for advancing computing efficiency and capabilities of AI applications using brain-inspired principles. However, the neuromorphic research field currently lacks standardized benchmarks, making it difficult to accurately measure technological advancements, compare performance with conventional methods, and identify promising future …

Towards a Bio-Inspired Real-Time Neuromorphic Cerebellum

Petrut Bogdan, Beatrice Marcinnò, Claudia Casellato, Stefano Casali et autres

This work presents the first simulation of a large-scale, bio-physically constrained cerebellum model performed on neuromorphic hardware. A model containing 97,000 neurons and 4.2 million synapses is simulated on the SpiNNaker neuromorphic system. Results are validated against a baseline simulation of the …

gb, it (code pays fourni par la source)

Towards a Bio-Inspired Real-Time Neuromorphic Cerebellum

Petrut Bogdan, Beatrice Marcinnò, Claudia Casellato, Stefano Casali et autres

From Frontiers via Jisc Publications Router

Applications - Doing Stuff on the Machine

Petrut Bogdan, Robert James, Gabriel Fonseca Guerra, Garibaldi Pineda García et autres

This chapter covers a wide range of topics. We begin by first presenting an art exhibit and SpiNNaker’s place in it – we start light. Then, we present a suite of approaches to engineer SNNs for a variety of computer vision tasks. …

4. Stacks of Software Stacks

Andrew Rowley, Oliver Rhodes, Petrut Bogdan, Christian Brenninkmeijer et autres

Learning in Neural Networks

Petrut Bogdan, Garibaldi Pineda García, Michael J. Hopkins, Edward Jones et autres

This chapter is concerned with the motivation, design and implementation behind mimicking biological learning rules with a focus on, you guessed it, SpiNNaker. It starts by presenting Spike-timing-dependent plasticity (STDP) operating in an unsupervised fashion based on relative spike times of the …

7. Learning in Neural Networks

Petrut Bogdan, Garibaldi Pineda García, Michael J. Hopkins, Edward Jones et autres

5. Applications - Doing Stuff on the Machine

Petrut Bogdan, Robert James, Gabriel Fonseca Guerra, Garibaldi Pineda García et autres

Event-based computation: unsupervised elementary motion decomposition

Petrut Bogdan, Garibaldi Pineda García, Simon Davidson, R. A. James et autres

Fast, localised motion detection is crucial for an efficient attention mechanism. We show that modelling a network capable of such motion detection can be performed using spiking neural networks simulated on many-core neuromorphic hard-ware. Moreover, highly sensitive neurons arise from the presented …

Supporting Data and Software for Event-based computation: Unsupervised elementary motion decomposition

Petrut Bogdan

We show that the presented architecture allows for unsupervised learning; that synaptic rewiring enhanced to initialise synapses by drawing from a distribution of delays produces more specialised neurons for the task of motion decomposition; and that a pair of readout neurons is …