The RWTH 2007 TC-STAR evaluation system for european English and Spanish
The RWTH 2007 TC-STAR evaluation system for european English and Spanish
复制标题
RWTH 2007 TC-STAR 欧洲英语和西班牙语评估系统
DOI：
--
复制
发表时间：
2007
期刊：
影响因子：
--
通讯作者：
H. Ney
中科院分区：
文献类型：
--
作者：
J. Lööf;Christian Gollan;Stefan Hahn;G. Heigold;Björn Hoffmeister;Christian Plahl;David Rybach;R. Schlüter;H. Ney
在这项工作中，RWTH自动语音识别系统开发的第三TC-STAR评估活动2007年。RWTH系统系统地使用内部系统组合，将系统与特征提取，自适应方法和所使用的训练数据的差异相结合。为了利用这一点，采用了新的特征提取方法;今年引入了Gammatone特征和基于MLP的电话后验特征。使用无监督训练实现了进一步的改进，值得注意的是，这些改进是使用相当少量的自动转录数据实现的。与去年相比，转向MPE培训以及引入投影SAT变换也有助于改进。
In this work, the RWTH automatic speech recognition systems developed for the third TC-STAR evaluation campaign 2007 are presented. The RWTH systems make systematic use of internal system combination, combining systems with differences in feature extraction, adaptation methods, and training data used. To take advantage of this, novel feature extraction methods were employed; this year saw the introduction of Gammatone features and MLP based phone posterior features. Further improvements were achieved using unsupervised training, and it is notable that these improvements were achieved using a fairly low amount of automatically transcribed data. Also contributing to the improvements over last year was the switch to MPE training, and the introduction of projecting SAT transforms.