WISARD is an early example of a hardware neural network pattern recognition machine (1981) (https://en.wikipedia.org/wiki/RAMnets#RAM-discriminators_and_WiSARD).
WISARD was designed and built by Bruce Wilkie at Brunel University as part of his Phd project with the backing of Professor Igor Aleksander and Dr John Stonham.
WISARD was a hardware machine that used a parallel architecture to run at video frame rates.
WISARD is on display at the Science Museum in the Winton Gallery (https://www.sciencemuseum.org.uk/learning/mathematics-winton-gallery-school-info).
This project is a software implementation of WISARD that performs the processing steps within your browser.
This implementation is written in Typescript as an Angular 7 service.
This implementation uses WebGL2 and the Shading Language to accelerate the processing pipeline.
This demonstration uses a SPA (https://en.wikipedia.org/wiki/Single-page_application) top-level application to walk you through WISARD's processing steps.
A runnable demo can be found at: https://malcolmbinstead.github.io/wisard/
And the source code can be found at: https://github.com/malcolmbinstead/wisard
Please note that this demo requires that you are using a browser and display card that supports webgl2.