首页> 外文会议>2nd Conference on Speech Technology and Human-Computer Dialogue; Apr 10-11, 2003; Bucharest >A COMPARATIVE STUDY OF THE METHODS USED IN ISOLATED WORDS RECOGNITION
【24h】

A COMPARATIVE STUDY OF THE METHODS USED IN ISOLATED WORDS RECOGNITION

机译:隔离词识别中使用方法的比较研究

获取原文
获取原文并翻译 | 示例

摘要

The present paperwork makes a comparative study of the methods used in isolated words recognition. It shows some experimental results obtained with the same word database, using three different methods: the first one based on Dynamic Time Warping (DTW), the second one based on Hidden Markov Models (HMM), and the last one based on Neural Networks. The conclusions point out that for a small database both DTW and HMM are good, but the ANN needs a larger database, to be trained with, in order to produce similar results.
机译:本文对孤立词识别中使用的方法进行了比较研究。它显示了使用三个不同方法在同一单词数据库中获得的一些实验结果:第一个基于动态时间规整(DTW),第二个基于隐马尔可夫模型(HMM),最后一个基于神经网络。结论指出,对于较小的数据库,DTW和HMM都不错,但是ANN需要使用较大的数据库进行训练,以便产生相似的结果。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号