Home

Welcome to the Neural Engineering Visual Database home!

MNIST spike representations

We provide Python code to convert existing computer vision databases into spiking representations; the conversion methods available are (NE15):

• Poisson
• Rank-ordered (via the FoCal algorithm)

Other sets of spiking representation included are:

• DVS recordings of flashing images (https://goo.gl/ru0fXP)
• DVS recordings of flashing images (http://www2.imse-cnm.csic.es/caviar/MNISTDVS.html)

Neuromorphic hardware

A list of reviewed hardware characteristics can be found here

Benchmark networks:

• Study Case I
• Study Case II