基于矢量泰勒级数的模型自适应算法.pdfVIP

第 32 卷第 1 期 电 子 与 信 息 学 报 Vol.32No.1 2010 年 1 月 Journal of Electronics Information Technology Jan. 2010 基于矢量泰勒级数的模型自适应算法 吕 勇 吴镇扬 (东南大学信息科学与工程学院 南京 210096) 摘 要：在实际环境中，由于测试环境与训练环境的不匹配，语音识别系统的性能会急剧恶化。模型自适应算法是 减小环境失配影响的有效方法之一，它通过测试环境下的少量自适应数据，将 HMM 模型的参数变换到测试环境 下。该文将矢量泰勒级数用于模型自适应，同时对 HMM 模型的均值向量和协方差矩阵进行变换，使其与实际环 境相匹配。实验证明，该文算法优于 MLLR 算法和基于矢量泰勒级数的特征补偿算法，在低信噪比环境中性能提 高尤为明显。 关键词：语音识别；模型自适应；矢量泰勒级数；隐马尔可夫模型 中图分类号：TN912.34 文献标识码：A 文章编号：1009-5896(2010)01-0107-05 DOI: 10.3724/SP.J.1146.2008.01768 Model Adaptation Algorithm Using Vector Taylor Series Lü Yong Wu Zhen-yang (School of Information Science and Engineering, Southeast University, Nanjing 210096, China) Abstract: In actual environments the performance of speech recognition system may be degraded significantly because of the mismatch between the training and testing conditions. Model adaptation is an efficient approach that could reduce this mismatch, which adapts model parameters to new conditions by some adaptation data. In this paper, a new model adaptation using vector Taylor series is presented, which adapts the mean vector and covariance matrix of hidden Markov model. The experimental results show that the proposed algorithm is more effective than MLLR and the feature compensation algorithm based on vector Taylor series in various environments, especially in low signal-to-noise ratio environments. Key words: Speech recognition; Model adaptation; Vector Taylor series; Hidden Markov model 1 引言 补偿算法恢复的特征向量误差较大，其性能明显低 在噪声环境下，由于训练环境和测试环境的失 于模型自适应算法。 配，语音识别系统的性能会急剧恶化