A Neural Net DTW Architecture in Speech Recognition

Li Haizhou; Xu Bingzheng

您当前的位置：

首页 >

文章列表页 >

A Neural Net DTW Architecture in Speech Recognition

更新时间：2025-12-08

- A Neural Net DTW Architecture in Speech Recognition
- Acta Electronica Sinica Issue 10, Pages: 82-87(1992)
- 作者机构：
  
  1. 华南理工大学
  2. 华南理工大学广州510641
- 作者简介：
- 基金信息：
- DOI：
  CLC：
- Published：1992
- 稿件说明：
移动端阅览
[1]李海洲,徐秉铮.语音识别中的神经网络DTW结构[J].电子学报,1992(10):82-87.

Li Haizhou, Xu Bingzheng. A Neural Net DTW Architecture in Speech Recognition[J]. Acta Electronica Sinica, 1992, (10): 82-87.
[1]李海洲,徐秉铮.语音识别中的神经网络DTW结构[J].电子学报,1992(10):82-87. DOI：

Li Haizhou, Xu Bingzheng. A Neural Net DTW Architecture in Speech Recognition[J]. Acta Electronica Sinica, 1992, (10): 82-87. DOI：

摘要

本文介绍语音识别中一种基于动态规划技术的时间规正算法DTW的神经网络实现方法.DTW是语音识别中最为有效的方法之一

它具有较强的鲁棒性且为语音识别系统提供了可能的最高识别率.但由于其计算量太大

除非用专门的硬件

DTW算法在实现时受到了限制.在本文中

所有的计算是由两个循环神经子网和一记忆层来完成的

该方法展示了算法的硬接线结构

（hard-wiring）的优越性

这为DTW的硬件实现提供了一种新的实施策略.

Abstract

This paper reports an implementation of dynamic programming based time-normalized algorithm

called dynamic time warping （DTW）

on neural networks. DTW is one of the most successful algorithms for spoken word recognition. It is very robust and usually provides the highest recognition rate possible but it takes a lot of computer time unless it is implemented by special hardware. In this implementation

the computation is governed by two recurrent subnets and one memory layer

demonstrating a hard-wiring mechanism which benefits from existing approaches.

关键词

Keywords

references

Views

117

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

A New Approach for Face Recognition Based on Singular Value Features and Neural Networks

Bionic Pattern Recognition—— A New Model of Pattern Recognition Theory and Its Applications (Lab of Artificial Neural Networks,Institute of Semiconductors,CAS,Beijing 100083,China)

A Neural Network That Can Recognize Some Common Abstract Features of Different Samples

Pattern Classification Neural Network Model Based on Homologue Connectedness

The Equal-Error Range Approximation and Shrink Learning Rule for Multilayer Perceptions and Its Applications

Related Author

GAN Jun-ying

ZHANG You-wei

WANG Shou-jue

LI Qian

DENG Hao-jiang

WANG Shou-jue

YANG Guo-wei

WANG Shou-jue

Related Institution

School of Information,Wuyi University

School of Electronics and Information Engineering, Beihang University

School of InformationWuyi UniversityJiangmenGuangdong 529020China

School of Electronics and Information EngineeringBeihang UniversityBeijing 100083China

Topological

⁰