一种适于低信噪比环境的语音识别智能玩具小车设计  

Design of a Speech Recognition Intelligent Toy Car Suitable for Low SNR Environment

作　　者：董胡 陈琦 彭高丰 陈耀东 刘刚 DONG Hu;CHEN Qi;PENG Gaofeng;CHEN Yaodong;LIU Gang(School of Information Science and Engineering,Changsha Normal University,Changsha 410100,China;School of Physics,Central South University,Changsha 410083,China)

机构地区：[1]长沙师范学院,信息科学与工程学院,湖南长沙410100 [2]中南大学,物理学院,湖南长沙410083

出　　处：《微型电脑应用》2024年第12期6-9,共4页Microcomputer Applications

基　　金：教育部人文社会科学研究青年基金项目(22YJCZH025);教育部产学合作协同育人项目(202101142015);湖南省教育厅重点研究项目(21A0618);长沙市自然科学基金项目(kq2202313)。

摘　　要：常见的语音识别智能玩具小车在高信噪比环境下表现良好,但在低信噪比环境下其语音识别性能表现较差,导致小车无法有效地执行相关动作。为此,提出一种适于低信噪比环境下的语音识别智能玩具小车设计。小车通过麦克风采集语音信号,利用深度学习算法及语音识别模块进行语音识别,主控芯片根据语音识别结果控制电机运转,实现控制玩具小车执行前进、后退、左转或右转动作。小车在行进过程中通过红外光电传感器进行避障。在低信噪比环境下,所设计的玩具小车语音识别正确率高,能正确实现既定动作,有较好的应用前景。Common speech recognition intelligent toy car performs well in high signal-to-noise ratio(SNR)environment,but their speech recognition performance is poor in low SNR environment,which leads to the fact that the car is unable to perform relevant actions effectively.To this end,this paper proposes a speech recognition intelligent toy car design suitable for low SNR environment.The car collects speech signal through microphone,uses deep learning algorithm and speech recognition module for speech recognition,and the main control chip controls the motor operation according to the speech recognition result to control the toy car to perform forward,backward,left or right turn actions.The car performs obstacle avoidance through infrared photoelectric sensor in the process of traveling.Under the low SNR environment,the designed toy car has a high speech recognition accuracy rate and can correctly realize the established actions,which has a better application prospect.

关 键 词：低信噪比 深度学习 语音识别 玩具小车 

分 类 号：TS958[轻工技术与工程]