/PP-LCNet

PyTorch implementation of PP-LCNet

Primary LanguagePythonMIT LicenseMIT

PyTorch implementation of PP-LCNet

Reproduction of PP-LCNet architecture as described in PP-LCNet: A Lightweight CPU Convolutional Neural Network by C. Cui. T. Gao, S. Wei et al (2021) with the PyTorch framework.

The official design is implemented with Paddle framework, the detail here

Models

Architecture #Parameters FLOPs Top-1 Acc. (%)
PPLCNet_x0_25 1,522,960 18M
PPLCNet_x0_35 1,646,888 29M
PPLCNet_x0_5 1,881,864 47M
PPLCNet_x0_75 2,359,792 99M
PPLCNet_x1_0 2,955,816 161M
PPLCNet_x1_5 4,504,136 342M
PPLCNet_x2_0 6,526,824 590M
PPLCNet_x2_5 9,023,880 906M

Stay tuned for ImageNet pre-trained weights.

Acknowledgement

The implementation is heavily borrowed from HBONet or MobileNetV2, please kindly consider citing the following

@InProceedings{Li_2019_ICCV,
author = {Li, Duo and Zhou, Aojun and Yao, Anbang},
title = {HBONet: Harmonious Bottleneck on Two Orthogonal Dimensions},
booktitle = {The IEEE International Conference on Computer Vision (ICCV)},
month = {Oct},
year = {2019}
}
@InProceedings{Sandler_2018_CVPR,
author = {Sandler, Mark and Howard, Andrew and Zhu, Menglong and Zhmoginov, Andrey and Chen, Liang-Chieh},
title = {MobileNetV2: Inverted Residuals and Linear Bottlenecks},
booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2018}
}