This Matlab toolbox implements some techniques for SPD matrices. The following techniques are available:
- SPD-COV: compute covariance on the SPD manifold
- SPD-LDV: compute Riemannian local difference vectors on the SPD manifold
- SPD-Means: compute Riemannian means on the SPD manifold
- SPD-Metrics: compute geodesic distances on the SPD manifold
Written by Kai-Xuan Chen (e-mail: chenkx.jsh@aliyun.com, chenkx@zju.edu.cn)
If you find this code useful for your research, we appreciate it very much if you can cite our related works:
BibTex :
@article{chen2020covariance,
title={Covariance Descriptors on a Gaussian Manifold and their Application to Image Set Classification},
author={Chen, Kai-Xuan and Ren, Jie-Yi and Wu, Xiao-Jun and Kittler, Josef},
journal={Pattern Recognition},
pages={107463},
year={2020},
publisher={Elsevier}
}
BibTex :
@inproceedings{chen2018riemannian,
title={Riemannian kernel based Nystr{\"o}m method for approximate infinite-dimensional covariance descriptors with application to image set classification},
author={Chen, Kai-Xuan and Wu, Xiao-Jun and Wang, Rui and Kittler, Josef},
booktitle={International conference on pattern recognition (ICPR)},
pages={651--656},
year={2018},
organization={IEEE}
}