Visual Media Processing

Visual Media Processing. Visual media has properties – high volume, high semantic and inter-class variety, high velocity – that enable us to push the boundaries of current AI modelling and software engineering techniques. It provides unique datasets that re quire AI methods that address feature reduction, multi-class issues, optimisation, semantic modelling and data analytics methods for managing big data, multi-source integration and privacy preservation. This produces a rich vein of topics to support PhD research.
Visual media is a valuable data source in domains with high potential to improve society, economic outcomes and knowledge. For example, smart cities, instrumented or autonomous vehicles, environmental sensing, health, and many others. The value of the CRT is that the student cohort will have access to supervisors with specific experience in advanced methods for visual media analytics and the network to facilitate transfer of methodologies between distinct but complementary application domains.