convert2snn 0.0.3

Convert conventional to spiking neural networks.

Convert2SNN

Tool for converting conventional neural networks to spiking neural networks.

Currently under construction. Please check again later until everything is implemented.

About

This tool can be used to convert and optimize conventional neural networks that were trained in TensorFlow to spiking neural networks (SNN).

Installation

Run the following to install:

pip install convert2snn

Usage

from convert2snn import convert

Developing

To install Convert2SNN alongside the tools you need to develop and run tests, run the following:

pip install -e.[dev]

References

• Mueller, Auge, Klimaschka, Knoll, "Neural Oscillations for Energy-Efficient Hardware Implementation of Sparsely Activated Deep Spiking Neural Networks", AAAI Practical DL, 2022
• Mueller, Studenyak, Auge, Knoll, "Spiking Transformer Networks: A Rate Coded Approach for Processing Sequential Data", ICSAI, 2021
• Mueller, Auge, Knoll, "Normalization Hyperparameter Search for Converted Spiking Neural Networks", Bernstein Conference, 2021
• Mueller, Hansjakob, Auge, Knoll, "Minimizing Inference Time: Optimization Methods for Converted Deep Spiking Neural Networks", IJCNN, 2021
• Mueller, Hansjakob, Auge, "Faster Conversion of Analog to Spiking Neural Networks by Error Centering", Bernstein Conference, 2020