This project implements an extended model of Spiking Neural P Systems (SN P systems) applied to image classification tasks. The framework supports both small example systems and a multilayer architecture designed for processing structured inputs such as images.
The project has been developed in collaboration between the University of Milano-Bicocca and the University of Verona.
The initial codebase is a fork of: https://github.com/a1sabau/spiking-p-system
The framework includes: - Extended SN P systems (multi-spike firing rules) - White hole mechanism - Inhibitory synapses with anti-spikes - Lightweight architecture without synaptic weights
It can be used to run small SN P system examples or a multilayer model for image classification, with the structure illustrated below.
Clone the repository:
git clone https://github.com/SandroErba/spiking-p-system.git
cd spiking-p-systemCreate a virtual environment (optional but recommended):
python -m venv venv
source venv/bin/activate # Linux / Mac
venv\Scripts\activate # WindowsInstall dependencies:
pip install -r requirements.txtRun the main simulation using the GUI:
python main_gui.pyIf a different entry point is used:
python main.pyExperiments can be executed by varying parameters in the GUI or directly in the code.
The model avoids the use of synaptic weights in order to remain as faithful as possible to the original P system framework. This results in a lightweight computational model, where performance is influenced primarily by structural design choices rather than parameter tuning.