|本期目录/Table of Contents|

[1]杨志杰,张 梅,李冠龙,等.基于长短时记忆元的语音智能识别系统设计[J].电子设计工程,2020,28(01):55-58.[doi:10.14022/j.issn1674-6236.2020.01.013]
 YANG Zhijie,ZHANG Mei,LI Guanlong,et al.Design of speech intelligent recognition system based on long?term and short?term memory elements[J].Electronic Design Engineering,2020,28(01):55-58.[doi:10.14022/j.issn1674-6236.2020.01.013]
点击复制

基于长短时记忆元的语音智能识别系统设计(PDF)
分享到:

《电子设计工程》[ISSN:1674-6236/CN:61-1477/TN]

卷:
28
期数:
2020年01期
页码:
55-58
栏目:
计算机技术应用
出版日期:
2020-01-05

文章信息/Info

Title:
Design of speech intelligent recognition system based on long?term and short?term memory elements
文章编号:
1674-6236(2020)01-0055-04
作者:
杨志杰张 梅李冠龙黄昌达
杨志杰,张 梅,李冠龙,黄昌达
Author(s):
YANG Zhi?jieZHANG MeiLI Guan?longHUANG Chang?da
(State Grid Xinjiang Electeic Power Research Institute CO.,LTD,Urumqi 830000,China)
关键词:
长短时记忆元 语音识别 智能识别 语音识别系统 声学模型
Keywords:
long-term and short-term memory element speech recognition intelligent recognition speech recognition system acoustic model
分类号:
TN912.34
DOI:
10.14022/j.issn1674-6236.2020.01.013
文献标志码:
A
摘要:
传统的语音智能识别系统对低资源语言的识别能力较差,识别结果准确性低。为解决这一问题,基于长短时记忆元设计了一种新的语音智能识别系统,针对系统硬件和软件进行优化。选用SPCE061A单片机作为系统硬件的核心处理器,并在硬件内部构建了LSTM—HMM声学模型,通过I/O接口连接各单元模块,无线传输模块内部核心芯片为nRF2401A芯片。软件程序由过构建声学模型、提取语言特征、监督训练语言文本、判断语速、语音识别优化等步骤完成,利用神经元网络设计了语音特征提取程序。为检测设计的识别系统工作效果,与传统识别系统进行实验对比,结果表明,相较于传统系统,基于长短时记忆元设计的语音智能识别系统对低资源词汇的识别错误率下降了29.3%,查询的关键词权重代价提升了71.2%。
Abstract:
Traditional speech intelligent recognition system has poor recognition ability for low-resource languages and low accuracy of recognition results. To solve this problem, a new intelligent speech recognition system based on long-term and short-term memory elements is designed, and the hardware and software of the system are optimized. SPCE061A MCU is selected as the core processor of the system hardware, and LSTM-HMM acoustic model is built in the hardware. Each unit module is connected through I/O interface. The core chip of the wireless transmission module is nRF2401A chip. The software program is completed by constructing acoustic model, extracting language features, supervising and training language text, judging speech speed, speech recognition optimization and so on. The speech feature extraction program is designed by using neural network structure. In order to test the performance of the recognition system, the experiment results show that compared with the traditional system, the recognition error rate of the speech intelligent recognition system based on long-term and short-term memory elements is reduced by 29.3%, and the cost of keyword weight is increased by 71.2%.

参考文献/References:

[1] RYAD Chellali. 基于双向长短时记忆-联结时序分类和加权有限状态转换器的端到端中文语音识别系统[J]. 计算机应用, 2018,38(9):2495-2499. [2] 舒帆,屈丹,张文林,等. 采用长短时记忆网络的低资源语音识别方法[J]. 西安交通大学学报, 2017, 51(10):120-127. [3] 姚增伟,刘炜煌,王梓豪,等. 基于卷积神经网络和长短时记忆神经网络的非特定人语音情感识别算法[J]. 新型工业化, 2018,45(9):1-8. [4] 葛瑞,王朝晖,徐鑫,等. 基于多层卷积神经网络特征和双向长短时记忆单元的行为识别(英文)[J]. 控制理论与应用, 2017,34(6):790-796. [5] 於帮兵,王华忠,颜秉勇. 基于长短时记忆网络的工业控制系统入侵检测[J]. 信息与控制, 2018,47(1):54-59.[6] 居治华,刘罡,陈琦岚,等. 基于反向卷积的Bi-LSTM语音识别[J]. 软件导刊,2018,17;189(7):31-34,40. [7] 李云红,王成,王延年. 基于混合DBNN-BLSTM模型的大词汇量连续语音识别[J]. 纺织高校基础科学学报, 2018,31(1):103-107. [8] 黄光许,田垚,康健,等. 低资源条件下基于i-vector特征的LSTM递归神经网络语音识别系统[J]. 计算机应用研究, 2017,34(2):392-396. [9] 冯伟,易绵竹,马延周. 基于TensorFlow的俄语词汇标音系统[J]. 计算机应用, 2017,38(4):971-977. [10]欧微,柳少军,贺筱媛,等. 战场对敌目标战术意图智能识别模型研究[J]. 计算机仿真, 2017,34(9):10-14. [11]郑毅,李凤,张丽,等. 基于长短时记忆网络的人体姿态检测方法[J]. 计算机应用, 2018,38(6):1568-1574. [12]刘畅,张一珂,张鹏远,等. 基于改进主题分布特征的神经网络语言模型[J]. 电子与信息学报, 2018,40(1):219-225.[13]司念文,王衡军,李伟,等. 基于注意力长短时记忆网络的中文词性标注模型[J]. 计算机科学, 2018,45(4):66-70.[14]成强. 基于小波包与长短时记忆融合的铁路旅客流量预测模型[J]. 计算机系统应用, 2018,27(7):123-128. [15]赵建鹏,周俊. 基于长短时记忆网络的旋转机械状态预测研究[J]. 噪声与振动控制, 2017,37(4):155-159. [16]李泽彬,姚有峰,张飞龙,等. 基于单片机的智能家居语音控制系统设计[J]. 电子设计工程, 2017,25(8):175-177.

相似文献/References:

[1]易圆皓.基于MatlabGUI语音识别推箱子的设计[J].电子设计工程,2016,(15):60.
 YI Yuan-hao.Design a game of speech recognition sokoban based on MatlabGUI[J].Electronic Design Engineering,2016,(01):60.
[2]李景富,周 鹏.基于嵌入式系统的智能车辆出入管理系统研究[J].电子设计工程,2015,(03):157.
 LI Jing-fu,ZHOU Peng.Research on the intelligent vehicle access management system based on embedded programme[J].Electronic Design Engineering,2015,(01):157.
[3]刘宝明,雷 虎,吴莎莎.便携雷达语音识别装置的电路实现[J].电子设计工程,2014,(23):8.
 LIU Bao-ming,LEI Hu,WU Sha-sha.On the realization of voice identification device for single radar[J].Electronic Design Engineering,2014,(01):8.
[4]孙慧芳,莫淳栋.基于STM32智能家居系统的设计与实现[J].电子设计工程,2014,(19):82.
 SUN Hui-fang,MO Chun-dong.STM32-based design and realization of intelligent home systems[J].Electronic Design Engineering,2014,(01):82.
[5]林向南,何岭松.基于Android的智能轮椅设计与实现[J].电子设计工程,2014,(19):14.
 LIN Xiang-nan,HE Ling-song.Design and implementation of intelligent wheelchair based on Android[J].Electronic Design Engineering,2014,(01):14.
[6]李清霞,魏文红. 基于视频摘要的智能视频识别系统的设计与实现[J].电子设计工程,2013,(16):65.
 LI Qing-xia,WEI Wen-hong. Design and implementation of video intelligent recognition system based on video abstraction[J].Electronic Design Engineering,2013,(01):65.
[7]袁伟伟,胡 军,刘万凤. 空管模拟训练中指令的语音识别与合成技术研究[J].电子设计工程,2013,(06):8.
 YUAN Wei-wei,HU Jun,LIU Wan-feng. Air traffic training simulation about command recognition and synthesis[J].Electronic Design Engineering,2013,(01):8.
[8]夏俊杰,杨 青,赵 刚. 语音识别组件转化为控件的方法[J].电子设计工程,2012,(14):59.
 XIA Jun-jie,YANG Qing,ZHAO Gang. The method from speech recognition component to ActiveX control[J].Electronic Design Engineering,2012,(01):59.
[9]王维强. ADSP-BF531在嵌入式语音识别系统中的应用[J].电子设计工程,2012,(12 ):186.
 WANG Wei-qiang. Application of ADSP-BF531 in embedded speech recognition system[J].Electronic Design Engineering,2012,(01):186.
[10]徐 莉,王志刚,高玺广,等.GPRS在智能家居监控系统中的应用[J].电子设计工程,2011,(16):105.
 XU Li,WANG Zhigang,GAO Xiguang,et al.Application of GPRS in intelligent home monitor[J].Electronic Design Engineering,2011,(01):105.

备注/Memo

备注/Memo:
收稿日期:2019-05-08 稿件编号:201905033基金项目:国家电网公司科技项目(PD71-17-008)作者简介:杨志杰(1970—),男,新疆吉木萨尔人,助理政工师。研究方向:电力营销服务。
更新日期/Last Update: 2019-12-30