Gaussian Mixture Model based voice conversion algorithms.