https://doi.org/10.1140/epjs/s11734-025-01512-3
Regular Article
Spike-timing dependent plasticity learning of small spiking neural network for image recognition
1
Saratov State University, 83 Astrakhanskaya Street, 410012, Saratov, Russia
2
Saratov Branch of Kotelnikov Institute of Radioengineering and Electronics of Russian Academy of Sciences, 38 Zelenaya Street, 410019, Saratov, Russia
Received:
6
January
2025
Accepted:
6
February
2025
Published online:
18
February
2025
To solve the problem of image recognition using a neural network, networks consisting of a large number of neurons are usually used. We have investigated the possibility of using a small neural network to recognize simple black and white images with added noise. The network under study consists of neurons, which are capable of generating spikes in response to external forcing. An unsupervised learning of our spiking neural network is based on the spike-timing dependent plasticity (STDP) method. The result of image recognition is studied depending on the number of neurons in the network, synaptic weights, STDP method parameters, and noise intensity in the images. Depending on the parameters of STDP learning, two different variants of the output layer dynamics are observed. In the first case, only one neuron in the output layer exhibits spiking activity, and different images cause spikes in different output neurons. In the second case, when an image is fed into the network, spikes are generated in a group of neurons in the output layer, and the combination of such firing neurons is unique for different images.
Copyright comment Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
© The Author(s), under exclusive licence to EDP Sciences, Springer-Verlag GmbH Germany, part of Springer Nature 2025
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
