PyTorch based implementation of Spiking Neural Network layers:
- SpikingDenseLayer
- SpikingConv1DLayer
- SpikingConv2DLayer
- SpikingConv3DLayer
- ReadoutLayer
including optional lateral and recurrent connections.
If you use this code, please consider citing: R. Zimmer, T. Pellegrini , S. F. Singh, and T. Masquelier. Technical report: supervised training of convolutional spiking neural networks with PyTorch. In https://arxiv.org/abs/1911.10124
This work is based on F. Zenke's tutorial on surrogate gradient learning in Spiking Neural Networks (https://github.com/fzenke/spytorch).
To run the speech_commands notebook:
- Download training data at http://download.tensorflow.org/data/speech_commands_v0.01.tar.gz and put them in a folder "data/speech_commands/train"
- Download testing data at http://download.tensorflow.org/data/speech_commands_test_set_v0.01.tar.gz and put them in a folder "data/speech_commands/test"