From Linear Prediction HMM to a New Combined Model for Speech Recognition

OU Zhi-jian; WANG Zuo-ying

您当前的位置：

首页 >

文章列表页 >

From Linear Prediction HMM to a New Combined Model for Speech Recognition

更新时间：2025-07-16

- From Linear Prediction HMM to a New Combined Model for Speech Recognition
- Acta Electronica Sinica Vol. 30, Issue 9, Pages: 1313-1316(2002)
- 作者机构：
  
  清华大学电子工程系,北京,100084
- 作者简介：
- 基金信息：
- DOI：
  CLC： TN912
- Published：2002
- 稿件说明：
移动端阅览
OU Zhi-jian, WANG Zuo-ying. From Linear Prediction HMM to a New Combined Model for Speech Recognition[J]. Acta Electronica Sinica, 2002, 30(9): 1313-1316.
DOI：

OU Zhi-jian, WANG Zuo-ying. From Linear Prediction HMM to a New Combined Model for Speech Recognition[J]. Acta Electronica Sinica, 2002, 30(9): 1313-1316. DOI：

摘要

线性预测HMM(Linear Prediction HMM

LPHMM)并没有象传统HMM那样引入状态输出独立同分布假设

但实用中识别性能并不佳.通过分析两种HMM的各自优劣

本文提出了一种新的语音识别的混合模型

将语音静态特性(基于传统HMM)和动态特性(基于LPHMM)分别描述又有机结合在一起

更为精确地刻划了真实的语音现象

同时又继承使系统的实现改动很小和较小的计算量.汉语大词汇量非特定人连续语音识别的实验表明

混合模型的识别性能显著好于LPHMM和传统HMM.理论上

本文还给出了LPHMM的一组闭式参数重估公式.

Abstract

Linear prediction (LP) HMM does not make the independent and identical distribution (IID) assumption as traditional HMM;however it often produces unsatisfactory results in practice.In this paper

a new combined model for speech recognition is proposed

based on a new analysis of both HMMs' modeling strengths and weaknesses.The new model works with LPHMM as the dynamic part and traditional IID-based HMM as the static part;in addition

easy implementation and low cost are preserved.Experiments on speaker-independent continuous speech recognition demonstrated that the combined model performed much better than both LPHMM and traditional HMM.Theoretically

a new closed-form parameter re-estimation formula is suggested for training LPHMM.

关键词

Keywords

references

Views

794

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Research on Polynomial-Fitting Speech-Trajectory Model in Chinese Continuous Speech Recognition

Interpolation Adaptation Algorithm Based on Gaussian Similarity Analysis

Research of Lung Cancer Recognition Based on Chest CT Images

Speaker-dependent Medium Vocabulary Continuous Speech Recognition Based on HMM and VQ

Related Author

WANG Zuo-ying

OU Zhi-jian

LU Da-jin

WANG Zuo-ying

LV Ping

HE Wei

JIANG Hui-yan

CAO Wen-ming

Related Institution

School of Software,Northeastern University

Research Institute of Intelligent Information SystemZhejiang University of TechnologyHangzhouZhejiang 310014China

Lab of Artificial Neural NetworksInstitute of SemiconductorsCASBeijing 100083China

Research Institute of Intelligent Information System,Zhejiang University of Technology

Lab of Artificial Neural Networks,Institute of Semiconductors,CAS

⁰