ChannelNets
Created by Hongyang Gao, Zhengyang Wang, and Shuiwang Ji at Texas A&M University.
Introduction
ChannelNets are compact and efficent CNN via Channel-wise convolutions. It has been accepted in NIPS2018.
Detailed information about ChannelNets is provided in [arXiv tech report] (https://arxiv.org/abs/1809.01330).
Citation
@article{gao2018channelnets,
title={ChannelNets: Compact and Efficient Convolutional Neural Networks via Channel-Wise Convolutions},
author={Hongyang Gao and Zhengyang Wang and Shuiwang Ji},
journal={arXiv preprint arXiv:1809.01330},
year={2018}
}
Results
Models | Top-1 | Params | FLOPs |
---|---|---|---|
GoogleNet | 0.698 | 6.8m | 1550m |
VGG16 | 0.715 | 128m | 15300m |
AlexNet | 0.572 | 60m | 720m |
SqueezeNet | 0.575 | 1.3m | 833m |
1.0 MobileNet | 0.706 | 4.2m | 569m |
ShuffleNet 2x | 0.709 | 5.3m | 524m |
ChannelNet-v1 | 0.705 | 3.7m | 407m |
Configure the network
All network hyperparameters are configured in main.py.