Service Humanoid Robotics: Review and Design of A Novel Bionic-Companionship Framework

At present, industrial robotics focused more on motion control and vision; whereas Humanoid Service Robotics (HSRs) are increasingly being investigated among researchers' and practitioners' field of speech interactions. The problematic and quality of human-robot interaction (HRI) has become one of the hot potatoes concerned in academia. This paper proposes a novel interactive framework suitable for HSRs. The proposed framework is grounded on the novel integration of Trevarthen Companionship Theory and neural image generation algorithm in computer vision. By integrating the image-to-natural interactivities generation, and communicate with the environment to better interact with the stakeholder, thereby changing from interaction to a bionic-companionship. In addition, the article also reviews the research of neural image generation algorithms and summarizes the application cases of the algorithm structure in the field of robotics from a critical perspective. We believe that the new interactive bionic-companionship framework can enable HSRs to further develop towards robot companions.