Get our free extension to see links to code for papers anywhere online!

Chrome logo  Add to Chrome

Firefox logo Add to Firefox


Structured Convolution Matrices for Energy-efficient Deep learning

Jun 08, 2016
Rathinakumar Appuswamy, Tapan Nayak, John Arthur, Steven Esser, Paul Merolla, Jeffrey Mckinstry, Timothy Melano, Myron Flickner, Dharmendra Modha


Share this with someone who'll enjoy it:


We derive a relationship between network representation in energy-efficient neuromorphic architectures and block Toplitz convolutional matrices. Inspired by this connection, we develop deep convolutional networks using a family of structured convolutional matrices and achieve state-of-the-art trade-off between energy efficiency and classification accuracy for well-known image recognition tasks. We also put forward a novel method to train binary convolutional networks by utilising an existing connection between noisy-rectified linear units and binary activations.



   Access Paper Source



Share this with someone who'll enjoy it: