Spike Template Matching Network implemented using the Lava neuromorphic computing framework.
The original code was developed in a private repository, originally authored by:
This package provides a framework for building Lava networks that perform template matching on spike sequences. This can be run as an independent network or as a part of larger network.
Check out the tutorial for a simple example. Currently, the module only supports fixed-pt CPU simulations of the network.
This package requires lava.
It also requires numpy and scipy.
There's no requirements file at the moment, since this module should work once lava is set up.
This package contains the network implementation.
It is split into two classes, TemplateMatchingNetwork and BinaryPredictionNetwork.
This package contains the configuration information for the network.
The configuration is interfaced through the TemplateMatchingNetworkConfig.
The function default_tmn_config() can be used to load the default configuration file, found under stmnet/configs/.
Contains functions for creating input and output modules for the network. Supports binary input processes and spike output processes.
Contains the lava processes and models for the custom neurons used in the network. The models are only available for fixed-point CPU implementations.