tjxtt / scm-snn

[IJCAI'23] Spike Count Maximization for Neuromorphic Vision Recognition

Spike Count Maximization for Neuromorphic Vision Recognition

The matlab code for the paper "Spike Count Maximization for Neuromorphic Vision Recognition".

Usage

Prerequisites

• Matlab 2018a

Example for Running

1. Download the example data and extract baseline_classifier.mat, train_data.mat, and test_data.mat into the example folder. The example data is the CIFAR10-DVS feature set. It was extracted from a PLIF-based Spiking ResNet-18 trained with MSE loss and PiecewiseLeakyReLU surrogate gradient.
2. Start Matlab, and then run the run.m script.

Citation

If our work is helpful to you, please kindly cite our paper as:

@inproceedings{tang2023scm,
  title={Spike Count Maximization for Neuromorphic Vision Recognition},
  author={Jianxiong Tang and Jian-Huang Lai and Xiaohua Xie and Lingxiao Yang},
  booktitle={IJCAI},
  year={2023}
}