Share Email Print

Proceedings Paper

Fast integer approximations in convolutional neural networks using layer-by-layer training
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

This paper explores method of layer-by-layer training for neural networks to train neural network, that use approximate calculations and/or low precision data types. Proposed method allows to improve recognition accuracy using standard training algorithms and tools. At the same time, it allows to speed up neural network calculations using fast-processed approximate calculations and compact data types. We consider 8-bit fixed-point arithmetic as the example of such approximation for image recognition problems. In the end, we show significant accuracy increase for considered approximation along with processing speedup.

Paper Details

Date Published: 17 March 2017
PDF: 5 pages
Proc. SPIE 10341, Ninth International Conference on Machine Vision (ICMV 2016), 103410Q (17 March 2017); doi: 10.1117/12.2268722
Show Author Affiliations
Dmitry Ilin, Smart Engines Ltd. (Russian Federation)
Elena Limonova, Moscow Institute of Physics and Technology (Russian Federation)
Vladimir Arlazarov, Moscow Institute of Physics and Technology (Russian Federation)
Dmitry Nikolaev, Institute for Information Transmission Problems (Russian Federation)

Published in SPIE Proceedings Vol. 10341:
Ninth International Conference on Machine Vision (ICMV 2016)
Antanas Verikas; Petia Radeva; Dmitry P. Nikolaev; Wei Zhang; Jianhong Zhou, Editor(s)

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?