This PhD project, in collaboration with the Nottingham Arts company Makers of Imaginary Worlds (MoIW), will investigate, develop, and implement (in a live environment with a real audience) adaptive robotic behaviours that can respond to changing audience states, such as excitement, boredom and attention using Computer Vision (CV) and Artificial Intelligence (AI) methods, to enhance and enrich the audience experience.
Research into personalised robotic systems is highly significant to the world of today, where Internet of Things (IoT) devices are commonplace in our homes, and AI robotics are becoming increasingly present in not only industrial applications but also everyday consumer products. In order to enable the production of higher quality, more socially enriching, and trustworthy systems that can be accessed and accepted by all members of society, it is imperative that we achieve a greater understanding of the robotic control and decision mechanisms and the ethical and moral constructs that help define them.
One of the governing aims of this project is to develop computer sensing solutions to evaluate human behaviours and emotions and investigate how an AI robotic system might perceive, interact, and respond to audiences and how that could benefit the audience experience or degrade it. A key area of research that directly relates to human behavioural analysis is contextual awareness in robotic systems. Enabling a system to detect human behaviour or emotion and understand why the human is displaying those behaviours or emotions is a crucial aspect of improving the autonomy of social robotics and human-robot relationships.