iCovDs: A MATLAB repository from chenchkx

This repository includes two representations for describing image sets.

CSPD: A simple but very effective framework to generate low-dimensional discriminative representation.
iCovDs: Also a more discriminative representation (it can be considered as an improved version of CSPD).

If you find this code useful for your research, we appreciate it very much if you can cite our related works:

BibTex :

@article{Chen2018Component,  
      title={Component SPD matrices: A low-dimensional discriminative data descriptor for image set classification},  
      author={Chen, Kai-Xuan and Wu, Xiao-Jun},  
      journal={Computational Visual Media},  
      volume={4},  
      number={3},  
      pages={245--252},  
      year={2018},  
      publisher={Springer}  
}

BibTex :

@inproceedings{chen2019more,
  title={More About Covariance Descriptors for Image Set Coding: Log-Euclidean Framework based Kernel Matrix Representation},
  author={Chen, Kai-Xuan and Wu, Xiao-Jun and Ren, Jie-Yi and Wang, Rui and Kittler, Josef},
  booktitle={Proceedings of the IEEE International Conference on Computer Vision Workshops},
  pages={0--0},
  year={2019}
}

Also, you can find more discriminative representations for image set classification at:

RieCovDs: https://github.com/Kai-Xuan/RiemannianCovDs
AidCovDs: https://github.com/Kai-Xuan/AidCovDs

chenchkx/iCovDs