The significance of cross-modal interaction in audio-visual quality perception
Andrew N. RIMELL, M. P. Hollier · 1999
To maintain reliable service provision, it has become necessary to automatically determine the perceived quality of service (PQOS) of communication systems in real time. This enables service providers to monitor their systems and to optimise their networks. BT has an established history of using auditory models in live telephone networks to assess network performance. Users are beginning to demand bi-modal information streams (audio-visual rather than audio only) and it has thus become necessary to develop PQOS measures for multi-modal systems. When assessing and modelling multi-modal systems, it is imperative that cross-modal interaction is taken into account, as a simple combination of mono-modal models will not give a robust perceived-quality performance metric. This paper describes work carried out at BT Labs into modelling audio-visual quality perception with a particular emphasis on cross-modal interaction. The work described includes subjective tests to provide data for the model.