Pith. sign in

REVIEW

Regularized Binary Network Training

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 1812.11800 v3 pith:UQHL6PIO submitted 2018-12-31 cs.LG cs.CV

Regularized Binary Network Training

classification cs.LG cs.CV
keywords binarytrainingmethodfunctionnetworksneuralregularizationactivation
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

There is a significant performance gap between Binary Neural Networks (BNNs) and floating point Deep Neural Networks (DNNs). We propose to improve the binary training method, by introducing a new regularization function that encourages training weights around binary values. In addition, we add trainable scaling factors to our regularization functions. Additionally, an improved approximation of the derivative of the sign activation function in the backward computation. These modifications are based on linear operations that are easily implementable into the binary training framework. Experimental results on ImageNet shows our method outperforms the traditional BNN method and XNOR-net.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.