示例代码-对称正定流形（Symmetric Positive Definite Manifold ，简称SPD流形）上距离计算常用的四种度量

 

SPD 流形即为由 SPD 矩阵所张成的空间，那么对于任意两个SPD矩阵 X, Y ，它们之间的距离通常可以由如下四种方式来计算：

 

1.仿射不变黎曼度量 (Affine Invariant Riemannian Metric):

$$d_A^2(X,Y)=||\log (X^{-\frac{1}{2}}Y{X}^{-\frac{1}{2}})|{|}_F^2$$

2. Stein散度 (Stein divergence):

$$d_S^2(X,Y)=\log \det (\frac{X+Y}{2})-\frac{1}{2}\log \det (XY)$$

3. Jeffrey散度 (Jeffrey divergence):

$$d{}_J^2(X,Y)=\frac{1}{2}Tr(X^{-1}Y)+\frac{1}{2}Tr(Y^{-1}X)-n$$

4.对数欧氏度量 (Log-Euclidean Metric):

$$d{}_L^2(X,Y)=\parallel \log \left(X\right)-\log \left(Y\right)\parallel {}_F^2$$

 
示例代码：
Github : https://github.com/Kai-Xuan/SPD-OPERATIONS/tree/master/SPD-Metrics/
百度云: https://pan.baidu.com/s/1hPvMsfafGTOnLVOVUXSIYQ 提取码：xv9k

相关推荐
1. SPD流形上均值计算: https://blog.csdn.net/u013515929/article/details/106632595/
2. SPD流形上协方差计算: https://blog.csdn.net/u013515929/article/details/106646995/
3. SPD流形上差分向量计算: https://blog.csdn.net/u013515929/article/details/106644056/

 

如果这个内容对于您的研究工作有帮助，我们将非常感激您可以引用我们的文章：[1,2].

---

[1]. Chen K X, Ren J Y, Wu X J, et al. Covariance Descriptors on a Gaussian Manifold and their Application to Image Set Classification[J]. Pattern Recognition, 2020, 107: 107463. [link]
[2]. Chen K X, Wu X J, Wang R, et al. Riemannian kernel based Nyström method for approximate infinite-dimensional covariance descriptors with application to image set classification[C]//International conference on pattern recognition (ICPR). 2018, 651-656. [link]