Collaborative Layer-Wise Discriminative Learning in Deep Neural Networks
Abstract
Intermediate features at different layers of a deep neural network are known to be discriminative for visual patterns of different complexities. However, most existing works ignore such cross-layer heterogeneities when classifying samples of different complexities. For example, if a training sample has already been correctly classified at a specific layer with high confidence, we argue that it is unnecessary to enforce rest layers to classify this sample correctly and a better strategy is to encourage those layers to focus on other samples.
Cite
Text
Jin et al. "Collaborative Layer-Wise Discriminative Learning in Deep Neural Networks." European Conference on Computer Vision, 2016. doi:10.1007/978-3-319-46478-7_45Markdown
[Jin et al. "Collaborative Layer-Wise Discriminative Learning in Deep Neural Networks." European Conference on Computer Vision, 2016.](https://mlanthology.org/eccv/2016/jin2016eccv-collaborative/) doi:10.1007/978-3-319-46478-7_45BibTeX
@inproceedings{jin2016eccv-collaborative,
title = {{Collaborative Layer-Wise Discriminative Learning in Deep Neural Networks}},
author = {Jin, Xiaojie and Chen, Yunpeng and Dong, Jian and Feng, Jiashi and Yan, Shuicheng},
booktitle = {European Conference on Computer Vision},
year = {2016},
pages = {733-749},
doi = {10.1007/978-3-319-46478-7_45},
url = {https://mlanthology.org/eccv/2016/jin2016eccv-collaborative/}
}