论文部分内容阅读
本文提出一种新的说话人自适应方法 :最大后验 (MAP)估计与最近邻线性回归 (NNLR)结合的自适应 ,利用模型近邻信息和MAP自适应结果 ,建立线性回归模型 ,对没有自适应数据的模型完成模型调整 .实验证明 ,NNLR要优于另一种用于MAP自适应框架的模型插值方法 :向量域平滑 (VFS) .
In this paper, we propose a new speaker adaptation method: the combination of maximum a posteriori (MAP) estimation and nearest neighbor linear regression (NNLR). Using the model neighborhood information and MAP adaptive result, we establish a linear regression model, Adapting the model of the data to complete the model adjustment.Experiments show that NNLR is superior to another model interpolation method used in the MAP adaptive framework: Vector Domain Smoothing (VFS).