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

Chrome logo Add to Chrome

Firefox logo Add to Firefox

Functional Hashing for Compressing Neural Networks

May 20, 2016
Lei Shi, Shikun Feng, ZhifanZhu

As the complexity of deep neural networks (DNNs) trend to grow to absorb the increasing sizes of data, memory and energy consumption has been receiving more and more attentions for industrial applications, especially on mobile devices. This paper presents a novel structure based on functional hashing to compress DNNs, namely FunHashNN. For each entry in a deep net, FunHashNN uses multiple low-cost hash functions to fetch values in the compression space, and then employs a small reconstruction network to recover that entry. The reconstruction network is plugged into the whole network and trained jointly. FunHashNN includes the recently proposed HashedNets as a degenerated case, and benefits from larger value capacity and less reconstruction loss. We further discuss extensions with dual space hashing and multi-hops. On several benchmark datasets, FunHashNN demonstrates high compression ratios with little loss on prediction accuracy.

* submitted to NIPS 2016 

Share this with someone who'll enjoy it:

   Access Paper Source

Share this with someone who'll enjoy it: