bindsnet
snntorch
bindsnet | snntorch | |
---|---|---|
1 | 2 | |
1,433 | 1,099 | |
1.7% | - | |
8.6 | 9.2 | |
4 days ago | 19 days ago | |
Python | Python | |
GNU Affero General Public License v3.0 | MIT License |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
bindsnet
snntorch
-
Bio inspired computer vision
Spiking Neural Networks (SNNs): neural networks that use spiking neurons (i.e. neurons that communicate using asynchronous binary spikes similarly to biological neurons) instead of artificial neurons. Apart from this particularity, SNNs can be organized in any kind of topology we all know, like CNNs, ViT, etc. There are tons of approaches to train SNNs, like bio-inspired learning rules (STDP, three factor rules, etc) or adaptations of backprop (which remains the SOTA in a lot of vision tasks). A good resource to begin with backprop-trained SNNs: https://snntorch.readthedocs.io/en/latest/ .
-
How to train brain-inspired spiking neural networks using lessons from deep learning. Interactive Colab notebook links in thread.
Github: https://github.com/jeshraghian/snntorch
What are some alternatives?
norse - Deep learning for spiking neural networks
spikingjelly - SpikingJelly is an open-source deep learning framework for Spiking Neural Network (SNN) based on PyTorch.
norse - Deep learning with spiking neural networks (SNNs) in PyTorch.
pytorch-forecasting - Time series forecasting with PyTorch
OpenWorm - Repository for the main Dockerfile with the OpenWorm software stack and project-wide issues
TorchGA - Train PyTorch Models using the Genetic Algorithm with PyGAD
Spiking-Neural-Network-SNN-with-PyTorch-where-Backpropagation-engenders-STDP - What about coding a Spiking Neural Network using an automatic differentiation framework? In SNNs, there is a time axis and the neural network sees data throughout time, and activation functions are instead spikes that are raised past a certain pre-activation threshold. Pre-activation values constantly fades if neurons aren't excited enough.
Kilosort - Fast spike sorting with drift correction for up to a thousand channels
Sophysics2D
pycox - Survival analysis with PyTorch