摘要
以导盲机器人作为研究载体,设计了一种RFID和非特定人声识别技术相结合的语音交互系统。采用125 kHz低频RFID系统将标签作为语音启动的触发信标,并设计其具体布局结构;同时应用LD3320中文语音识别芯片,通过非特定人声搜索引擎和特征库计算出识别结果。机器人样机语音交互系统可完成对用户和环境状态信息的语音问答。不同语速、关键词的延时时间和识别率的测试和统计表明,该系统可以满足语音交互的基本需求。
With the guide robot for blind as the research carrier, the voice interaction system based on the combination of RFID and speaker-independent automatic speech recognition （ SI-ASR） technology has been designed. By using 125 kHz LF RFID system, and the labels are adopted as the trigger beacon to start up the speech, specific layout structure is designed;in addition, the LD3320 chinese speech recognition chip is applied, through speaker-independent voice searching engine and the feature database, the recognition results are calculated. The robot voice interaction system can complete the recognition function for user command and reply environment status information. Through the tests and statistics under different speech speed, time delay of the keywords and recognition rate, it is indicated that the system satisfies the basic requirement of voice interaction.
出处
《自动化仪表》
CAS
北大核心
2014年第3期73-76,共4页
Process Automation Instrumentation
基金
江苏省高校自然科学研究基金资助项目(编号:09KJD46003)
江苏省研究生创新计划基金资助项目(编号:CXLX12-0687)
