Abstract
We propose a method for interpolation of a high-dimensional rotation matrix and also demonstrate its applications in computer graphics domain and pattern recognition domain, which are interpolation of object poses in high-dimensional space and interpolation of multivariate normal distributions.