Complexity-aware loss function for fast neural networks with early exits

Lev Teplyakov; Segrey Gladilin; Evgeny Shvets

doi:10.1117/12.2557077

31 January 2020 Complexity-aware loss function for fast neural networks with early exits

Lev Teplyakov, Segrey Gladilin, Evgeny Shvets

Proceedings Volume 11433, Twelfth International Conference on Machine Vision (ICMV 2019); 114333I (2020) https://doi.org/10.1117/12.2557077
Event: Twelfth International Conference on Machine Vision, 2019, Amsterdam, Netherlands

Abstract

Most modern convolutional neural networks (CNNs) are compute-intensive, making them infeasible to use in mobile or embedded devices. One of the approaches to this problem is to modify a usual deep CNN with shallow early-exit branches, appended to some convolutional layers [1]. This modification, named BranchyNet, allows to process simple input samples without performing full volume of calculations, providing a speed-up on average. In this work we consider the problem of training a BranchyNet. We exploit a cascade loss function [2], which explicitly regularizes CNN’s average computation time, and modify it to use the entropy of branches’ prediction as confidence measure. We show, that on CIFAR10 dataset the proposed loss function provides a actual speed-up increase from 43% to 47% without quality degradation, comparing with the original loss function.

Citation Download Citation

Lev Teplyakov, Segrey Gladilin, and Evgeny Shvets "Complexity-aware loss function for fast neural networks with early exits", Proc. SPIE 11433, Twelfth International Conference on Machine Vision (ICMV 2019), 114333I (31 January 2020); https://doi.org/10.1117/12.2557077

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available