Robot Control Based On Hybrid Speech Interaction System

At present,the hotel service industry has a shortage of professional talents,high labor cost and large liquidity,so the hotel service robot will have a broad application prospect and a huge market value in the future.Based on the friendly arm board Tiny4412 with Android system,this paper studies the speech interaction system with low error rate and high real-time performance on the hotel service-robot in order to make the service-robot play a greater role in hotel management.The hotel will save costs and serve hotel guests better than before.The system includes: face detection,speech synthesis,speech understanding,speech recognition,building a database,lower computer communicationand and so on.OpenCV ported to the Android platform.The system uses the AdaBoost algorithm to achieve real-time face detection combined with the OpenCV library.The speech recognition chip HBR740 achieves offline speech recognition.The speech synthesizer chip SYN6288 achieves offline speech synthesis.The system obtaines a large amount of voice resources and achieves online speech recognition,online speech understanding,online speech recognition through the network connection to the cloud platform of IFLYTEK.speech recognition is applied to the motion control of the lower machine through Tiny4412's serial port to transmit data.The SQLite database of voice interaction and control movement is built through jar package of LitePal and POI.The local voice interaction and the online voice interaction are combined together.A hybrid voice interaction system is designed and tested to verify the feasibility.Under the condition of good network condition,the speech recognition rate of the whole system is 94.67%.In the absence of network,the recognition rate of speech recognition speech recognition reaches 84.67%.The system can detect the arrival of hotel guests through the face detection function,complete the natural voice interaction with the hotel guests and control the movement of the robot.