Efficient Approximation of Filters for High-Accuracy Binary Convolutional Neural Networks
Abstract
In this paper, we propose an efficient design to convert full-precision convolutional networks into binary neural networks. Our method approximates a full-precision convolutional filter by sum of binary filters with multiplicative and additive scaling factors. We present closed form solutions to the proposed methods. We perform experiments on binary neural networks with binary activations and pre-trained neural networks with full-precision activations. The results show an increase in accuracy compared to previous binary neural networks. Furthermore, to reduce the complexity, we prune scaling factors considering the accuracy. We show that up to a certain degree of threshold, we can prune scaling factors while maintaining accuracy comparable to full-precision convolutional neural networks.
Cite
Text
Park et al. "Efficient Approximation of Filters for High-Accuracy Binary Convolutional Neural Networks." European Conference on Computer Vision Workshops, 2020. doi:10.1007/978-3-030-68238-5_6Markdown
[Park et al. "Efficient Approximation of Filters for High-Accuracy Binary Convolutional Neural Networks." European Conference on Computer Vision Workshops, 2020.](https://mlanthology.org/eccvw/2020/park2020eccvw-efficient/) doi:10.1007/978-3-030-68238-5_6BibTeX
@inproceedings{park2020eccvw-efficient,
title = {{Efficient Approximation of Filters for High-Accuracy Binary Convolutional Neural Networks}},
author = {Park, Junyong and Moon, Yong-Hyuk and Lee, Yong-Ju},
booktitle = {European Conference on Computer Vision Workshops},
year = {2020},
pages = {71-84},
doi = {10.1007/978-3-030-68238-5_6},
url = {https://mlanthology.org/eccvw/2020/park2020eccvw-efficient/}
}