Anna Trofimova has presented a paper “Exploring Human-Human Interaction for Applications in Social HRI” at the 31 Australian Conference on Human-Computer Interaction (OZCHI 2019) in Fremantle, Australia. Co-authored with Anna’s supervisors Timothy Wiley, Maurice Pagnucco and Mari Velonaki, this paper describes the methodology and the approach undertaken to the collection of emotional expressions that will be used in Anna’s PhD research project to train a behaviour generation system for social robots. If you are interested in the paper you can find it here: Exploring Human-Human Interaction for Applications in Social HRI
The objective of this study is to develop a robot behavioural model that includes non-verbal vocal information to produce behaviour that fits user expectation better. Our approach is to train the behavioural model using data representing human-human interaction during collaboration. The data, voice and user control inputs, are gathered during a game where players collaborate to navigate a character through a path. This paper presents the details of the approach undertaken to the data collection procedure as well as preliminary analysis of the data.