Noise Robust Phonetic Classificationwith Linear Regularized Least Squares and Second-Order Features
Noise Robust Phonetic Classificationwith Linear Regularized Least Squares and Second-Order Features
复制标题
具有线性正则化最小二乘和二阶特征的噪声鲁棒语音分类
DOI：
--
复制
发表时间：
2007
期刊：
影响因子：
--
通讯作者：
James R. Glass
中科院分区：
文献类型：
--
作者：
R. Rifkin;K. Schutte;Michelle Saad;J. Bouvrie;James R. Glass
我们执行语音分类的架构，其元素是通过线性正则化最小二乘（RLS）训练的二进制分类器。RLS是一种简单但功能强大的正则化算法，具有理想的特性，即可以通过最小化训练集上的留一法误差来有效地找到正则化参数的良好值。我们的系统在TIMIT语音分类任务上实现了最先进的单分类器性能，（略微）击败了其他最近的系统。我们还表明，在存在加性噪声的情况下，我们的模型比训练良好的高斯混合模型更鲁棒。
We perform phonetic classification with an architecture whose elements are binary classifiers trained via linear regularized least squares (RLS). RLS is a simple yet powerful regularization algorithm with the desirable property that a good value of the regularization parameter can be found efficiently by minimizing leave-one-out error on the training set. Our system achieves state-of-the-art single classifier performance on the TIMIT phonetic classification task, (slightly) beating other recent systems. We also show that in the presence of additive noise, our model is much more robust than a well-trained Gaussian mixture model.