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

facebooktwittermailshare

An Efficient Alternative to Network Pruning through Ensemble Learning

Abstract: 

Convolutional Neural Networks (CNNs) currently represent the best tool for classification of image content. CNNs are trained in order to develop generalized expressions in form of unique features to distinguish different classes. During this process, one or more filter weights might develop the same or similar values. In this case, the redundant filters can be pruned without damaging accuracy.Unlike normal pruning methods, we investigate the possibility of replacing a full-sized convolutional neural network with an ensemble of its narrow versions. Empirically, we show that the combination two narrow networks, which has only half of the original parameter count in total, is able to contest and even outperform the full-sized original network. In other words, a pruning rate of 50% can be achieved without any compromise on accuracy. Furthermore, we introduce a novel approach on ensemble learning called ComboNet, which increases the accuracy of the ensemble even further.

up
0 users have voted:

Paper Details

Authors:
Martin Poellot, Rui Zhang, André Kaup
Submitted On:
29 May 2020 - 8:29am
Short Link:
Type:
Presentation Slides
Event:
Presenter's Name:
Martin Pöllot
Paper Code:
2638
Document Year:
2020
Cite

Document Files

20.04.13_ComboNet_16x9.pdf

(59)

Subscribe

[1] Martin Poellot, Rui Zhang, André Kaup, "An Efficient Alternative to Network Pruning through Ensemble Learning", IEEE SigPort, 2020. [Online]. Available: http://sigport.org/5447. Accessed: Dec. 02, 2020.
@article{5447-20,
url = {http://sigport.org/5447},
author = {Martin Poellot; Rui Zhang; André Kaup },
publisher = {IEEE SigPort},
title = {An Efficient Alternative to Network Pruning through Ensemble Learning},
year = {2020} }
TY - EJOUR
T1 - An Efficient Alternative to Network Pruning through Ensemble Learning
AU - Martin Poellot; Rui Zhang; André Kaup
PY - 2020
PB - IEEE SigPort
UR - http://sigport.org/5447
ER -
Martin Poellot, Rui Zhang, André Kaup. (2020). An Efficient Alternative to Network Pruning through Ensemble Learning. IEEE SigPort. http://sigport.org/5447
Martin Poellot, Rui Zhang, André Kaup, 2020. An Efficient Alternative to Network Pruning through Ensemble Learning. Available at: http://sigport.org/5447.
Martin Poellot, Rui Zhang, André Kaup. (2020). "An Efficient Alternative to Network Pruning through Ensemble Learning." Web.
1. Martin Poellot, Rui Zhang, André Kaup. An Efficient Alternative to Network Pruning through Ensemble Learning [Internet]. IEEE SigPort; 2020. Available from : http://sigport.org/5447