Sorry, you need to enable JavaScript to visit this website.

facebooktwittermailshare

TRAINING ACCURATE BINARY NEURAL NETWORKS FROM SCRATCH

Abstract: 

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

up
0 users have voted:

Paper Details

Authors:
Joseph Bethge, Haojin Yang, Christoph Meinel
Submitted On:
11 September 2019 - 8:14am
Short Link:
Type:
Poster
Event:
Presenter's Name:
Joseph Bethge
Paper Code:
MP.PB.4
Document Year:
2019
Cite

Document Files

poster.pdf

(18)

Subscribe

[1] Joseph Bethge, Haojin Yang, Christoph Meinel, "TRAINING ACCURATE BINARY NEURAL NETWORKS FROM SCRATCH", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4590. Accessed: Nov. 12, 2019.
@article{4590-19,
url = {http://sigport.org/4590},
author = {Joseph Bethge; Haojin Yang; Christoph Meinel },
publisher = {IEEE SigPort},
title = {TRAINING ACCURATE BINARY NEURAL NETWORKS FROM SCRATCH},
year = {2019} }
TY - EJOUR
T1 - TRAINING ACCURATE BINARY NEURAL NETWORKS FROM SCRATCH
AU - Joseph Bethge; Haojin Yang; Christoph Meinel
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4590
ER -
Joseph Bethge, Haojin Yang, Christoph Meinel. (2019). TRAINING ACCURATE BINARY NEURAL NETWORKS FROM SCRATCH. IEEE SigPort. http://sigport.org/4590
Joseph Bethge, Haojin Yang, Christoph Meinel, 2019. TRAINING ACCURATE BINARY NEURAL NETWORKS FROM SCRATCH. Available at: http://sigport.org/4590.
Joseph Bethge, Haojin Yang, Christoph Meinel. (2019). "TRAINING ACCURATE BINARY NEURAL NETWORKS FROM SCRATCH." Web.
1. Joseph Bethge, Haojin Yang, Christoph Meinel. TRAINING ACCURATE BINARY NEURAL NETWORKS FROM SCRATCH [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4590