项目作者: jiankang1991

项目描述 :
codes for RS paper: High-Rankness Regularized Semi-supervised Deep Metric Learning for Remote Sensing Imagery
高级语言: Python
项目地址: git://github.com/jiankang1991/HR-S2DML.git
创建时间: 2020-07-22T19:55:58Z
项目社区:https://github.com/jiankang1991/HR-S2DML

开源协议:

下载


High-Rankness Regularized Semi-supervised Deep Metric Learning for Remote Sensing Imagery

Jian Kang, Ruben Fernandez-Beltran, Zhen Ye, Xiaohua Tong, Pedram Ghamisi, Antonio Plaza.


This repo contains the codes for the MDPI RS paper: High-Rankness Regularized Semi-supervised Deep Metric Learning for Remote Sensing Imagery. we reformulate the deep metric learning scheme in a semi-supervised manner to effectively characterize RS scenes. Specifically, we aim at learning metric spaces by utilizing the supervised information from a small number of labeled RS images and exploring the potential decision boundaries for massive sets of unlabeled aerial scenes. In order to reach this goal, a joint loss function, composed of a normalized softmax loss with margin and a high-rankness regularization term, is proposed, as well as its corresponding optimization algorithm. Some codes are modified from ArcFace and BNM.

alt text

Usage

./train_MNCE_BNM/main.py is the script of the proposed method for training and validation.

Citation

  1. @article{kang2020highrank,
  2. title={{High-Rankness Regularized Semi-supervised Deep Metric Learning for Remote Sensing Imagery}},
  3. author={Kang, Jian and Fernandez-Beltran, Ruben and Ye, Zhen and Tong, Xiaohua and Ghamisi, Pedram and Plaza, Antonio},
  4. journal={Remote Sensing},
  5. year={2020},
  6. note={DOI:}
  7. publisher={MDPI}
  8. }

References

[1] Deng, Jiankang, et al. “Arcface: Additive angular margin loss for deep face recognition.” Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2019.

[2] Cui, Shuhao, et al. “Towards discriminability and diversity: Batch nuclear-norm maximization under label insufficient situations.” Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2020.