Communication Effiency and User Experience Analysis of Visual and Audio Feedback Cues in Human and Service Robot Voice Interaction Cycle
Yi. Xing,G. Hou,J. Tan,Haonan Song
DOI: https://doi.org/10.1109/WRC-SARA.2019.8931967
2019-08-01
Abstract:As the most natural way of interaction, voice interaction plays a huge role in the development and progress of home service robots. Currently, most studies research voice interaction from a technical perspective. When the accuracy of speech recognition is as high as 98%, the effect of voice interaction is still not optimistic. One reason is that there is a lack of discussion of interaction cycle design from a user experience perspective, ignoring the importance of cues during interaction. The home service robot discussed in this paper uses the Turtlebot as a platform to explore the importance and necessity of different feedback cues for awakening, listening, speaking and thinking in the cycle of voice interaction. In the paper, we used the feedback cues as a variable, we designed five test systems. The experimental results show that the cues for listening are the most important and necessary for the interaction cycle. When a system removes the cues of listening, it will bring huge obstacles and negative effects to the communication process. The wake-up cues are important and necessary. When there are not wake-up cues in the system, the task completion is greatly reduced. On the contrary, the cues of thinking, which are thought to be more important, can be used to indicate that the robot has received the voice message. However, according to the experimental data, its lack did not bring much hindrance to the experiment. Removing the cues of speaking does not have a negative impact on the experiment, but rather reduces the experimental time to a certain extent.
Computer Science,Engineering