Spiking Neural Network Implementation on FPGA

This is an FPGA imlementation of Spiking Neural Network Algorithm that supports simple image classification example.

Built With

• scikit-learn
• Numpy
• Matplotlib
• spikingjelly

Getting Started

Prerequisites

• Python version 3.9.0 >

  https://www.python.org/

• Virtualenv

  https://pypi.org/project/virtualenv/

Installation

1. Clone the repository

   git clone https://github.com/USTC-SSE-practice/SNN_FPGA.git

2. Create Virtual Environment

   virtualenv venv

   OR

   python -m venv venv

3. Activate Virtual Environment

   source venv\scripts\activate ~ on~windows

   source venv/bin/. activate ~ on~mac

4. Install Dependencies

   pip install -r requirements.txt

Running

• cd into the root where the project is cloned
• run src/inference.py
• this command will use pregenerated weights and output the results of classifying the test images

Weight Reconstruction (Training)

• to reconstruct the weights, run src/train.py
• this command will generate new or update the model_weights.pth file to be used for prediction in the inference.py

Notes

• in /hls folder only snn_tp.cpp and host2.cpp are required to upload to Vitis IDE