Design And Implementation Of Intelligent Voice Interaction System Based On ROS

Voice,as one of the most natural man-machine interaction modes,has achieved rapid development driven by artificial intelligence technology in recent years.Nowadays,voice interaction technology is changing people's living habits in various fields.The popularity of smart phones is an opportunity for voice-related products to enter millions of households.Voice personal assistants such as Apple's Siri have brought voice interaction technology to new heights of development;now smart speakers are like new comers in home entertainment;at the same time,there are endless voice education robots;in a sense,voice interaction is quietly emerging.The use of a voice cloud platform to build a product with voice interaction capabilities under the cloud technology architecture is currently the mainstream solution.The terminal device only needs to take charge of voice signal acquisition and final audio output.The rest of the work such as voice recognition and semantic understanding are all processed by voice service platform.The features of this program are: The main R&D effort is transferred to the terminal voice signal processing,and the intelligent decision-making is highly dependent on the cloud voice service platform.Although the products that currently use voice interaction as a selling point are in various forms,the product technology implementation plans share the same goal.Therefore,there are more or less duplicative design tasks in system construction,which will increase the development cycle and cost in the product development stage,these are not conducive to long-term development.The emergence of ROS provides a solution for maximizing software reusability.It is a distributed software design framework that can divide different functional modules into nodes and then adjust the functional differences by adjusting the communication links between the nodes.For the problems in the voice interaction system constructed under the cloud architecture,This article will use the software features of ROS to adjust the framework of the voice interaction system under the current cloud architecture.The main work includes the following aspects:(1)Investigate the implementation of the traditional speech interaction system,carding system to achieve the key technical points.Learning the ROS framework,and through the ROS software design ideas to adjust the traditional voice interactive system framework to increase the system scalability and maintainability.(2)Using the idea of multi-feature fusion and scanning to improve the traditional speech endpoint detection algorithm to improve the accuracy of the system in the endpoint detection.At the same time,in order to compensate for the misjudgment in this link,the loss of speech frames will affect the speech recognition process,and the speech frame buffer is especially designed.(3)In order to reduce the dependence of the terminal on the cloud semantic understanding service,an offline intention identification model is specially designed.The main purpose is to provide preconditions for collaborative scheduling of online and offline resources in order to improve the responsiveness of the system and the fluency of the interaction process.