通过一些额外的语音对训练好的说话人模型进行测试统计出其输出评分的均值和方差（u,δ）当真正进行测试时，对输出模型的分数score，用规整之后的score =（score-u）/δ代替。使得结果更加合理。-Speaker model trained by some additional voice test statistics of the mean and variance of its output rated (u, δ) when the real test, the fraction of the output model score with regularity after the score ' = (score-u)/δ instead. Makes the results more reasonable.