Training Accurate Binary Neural Networks from Scratch

Binary neural networks are a promising approach to execute convolutional neural networks on devices with low computational power. Previous work on this subject often quantizes pretrained full-precision models and uses complex training strategies. In our work, we focus on increasing the performance o...

Full description

Saved in:
Bibliographic Details
Published inProceedings - International Conference on Image Processing pp. 899 - 903
Main Authors Bethge, Joseph, Yang, Haojin, Meinel, Christoph
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.09.2019
Subjects
Binary Neural Networks
Computational modeling
Computer architecture
Convolution
Network architecture
Neural networks
Training
Online AccessGet full text

Cover

More Information
Summary:Binary neural networks are a promising approach to execute convolutional neural networks on devices with low computational power. Previous work on this subject often quantizes pretrained full-precision models and uses complex training strategies. In our work, we focus on increasing the performance of binary neural networks by training from scratch with a simple training strategy. In our experiments we show that we are able to achieve state-of-the-art results on standard benchmark datasets. Further, we analyze how full-precision network structures can be adapted for efficient binary networks and adopt a network architecture based on a DenseNet for binary networks, which lets us improve the state-of-the-art even further. Our source code can be found online: https://github.com/hpi-xnor/BMXNet-v2.
ISSN:2381-8549
DOI:10.1109/ICIP.2019.8802610