rar

Real-time implementation of continuous digital speech recognition system using fixed-point DSP

  • 2013-07-01
  • 224.1KB
  • Points it Requires : 1

For various continuous digital speech signals, a continuous digital speech recognition system based on TMS320C5x evaluation module (EVM) and fixed-point digital signal processor ADSP2181 is implemented. Based on the analysis of continuous probability density hidden Markov model (CDHMM), LPC cepstral coefficients, LPC difference cepstral coefficients, energy normalization coefficients and their difference coefficients are used as speech feature vectors. Viterbi algorithm and Baum-Welch re-estimation method are used for training and recognition, and ADSP2181 is used to implement the speech recognition algorithm. The recognition rate of the system is effectively improved. The time required for implementing each stage is given, and the influence of different speech feature parameters on the recognition rate is compared. In the specific implementation, the issues of noise resistance, fixed-point real-time implementation and identification of people by continuous digital strings are focused on. The experimental results show that the system achieves satisfactory results in ordinary environments, with a correct recognition rate of 93.2%, which provides an important technical approach for its practical application.

unfold

You Might Like

Uploader
论文帝
 

Recommended ContentMore

Popular Components

Just Take a LookMore

EEWorld
subscription
account

EEWorld
service
account

Automotive
development
circle

About Us Customer Service Contact Information Datasheet Sitemap LatestNews


Room 1530, 15th Floor, Building B, No.18 Zhongguancun Street, Haidian District, Beijing, Postal Code: 100190 China Telephone: 008610 8235 0740

Copyright © 2005-2024 EEWORLD.com.cn, Inc. All rights reserved 京ICP证060456号 京ICP备10001474号-1 电信业务审批[2006]字第258号函 京公网安备 11010802033920号
×