Speaker identification using minimum classification error training
Speaker identification using minimum classification error training
复制标题
使用最小分类误差训练进行说话人识别
DOI：
10.1109/icassp.1998.674379
复制
发表时间：
1998
期刊：
影响因子：
--
通讯作者：
S. Parthasarathy
中科院分区：
文献类型：
--
作者：
Olivier Siohan;A. Rosenberg;S. Parthasarathy
我们使用最小分类误差(MCE)训练范式来构建说话人识别系统。对于依赖于文本的说话人识别任务，在字符串级对训练进行优化。在小集合说话人识别任务上的实验表明，MCE训练可以比使用最大似然估计训练的基线系统减少高达20%-25%的闭集识别错误。进一步的实验表明，通过使用来自登记说话人集合之外的说话人的一些额外的训练数据，可以获得进一步的改进，导致闭合识别错误总体减少约35%。
We use a minimum classification error (MCE) training paradigm to build a speaker identification system. The training is optimized at the string level for a text-dependent speaker identification task. Experiments performed on a small set speaker identification task show that MCE training can reduce closed-set identification errors by up to 20-25% over a baseline system trained using maximum likelihood estimation. Further experiments suggest that additional improvement can be obtained by using some additional training data from speakers outside the set of registered speakers, leading to an overall reduction of the closed-set identification errors by about 35%.