Skip to content
Shield board with a trainable and scalable neural network
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.
FPGA configuration


NeuroMem shield board compatible with Arduino with 1 NM500 chips totaling 576 neurons ready to learn and recognize patterns extracted from your signal, images, measurements and other data source. Add a NeuroBrick to expand the network seamlessly with 1152 neurons.

Arduino library & examples for Arduino boards

Python library & examples for Raspberry Pi

  • Academic script to understand how to teach and query the neurons Video recognition example using RaspiCam

GV_NeuroMem API for USB interface

  • C++ library to access the neurons through the NeuroShield USB-serial port
  • Academic script to understand how to teach and query the neurons Note that this API is delivered for Windows. Adapting its souyrce code to Linux should just involve linking to the Cypress USB driver for Linux.

Interface to other hardware with SPI port

The source code of the primitive functions SPI_Connect, SPI_Read and SPI_Write can be found in:

  • Arduino\Libraries\Src\NeuroMemSPI.cpp
  • USB\NeuroMemAPI\lib\comm_neuroshield
  • Python ex\

The NeuroMem USB dongle is compatible with the following tools from General Vision:

  • NeuroMem Knowledge Builder for Training and Validation of a NeuroMem network on your datasets
  • CogniPat SDK for generic pattern learning and recognition, with examples in C++, C#, Python, and MatLab
  • CogniPat SDK for MatLab for generic pattern learning and recognition including examples on sound and image files
  • CogniPat SDK for LabVIEW for generic pattern learning and recognition in LabVIEW

To read more, visit

Hardware Specifications

For more details, refer to the nepes NeuroShield Hardware Manual at

You can’t perform that action at this time.