Keynote talks

Bio:

Rosie Campbell is a Technologist at BBC R&D. Her current project aims to explore how we can use machine learning techniques to create artificially intelligent production systems for live events. She has also been working towards a doctorate focusing on Convolutional Neural Networks for Visual Computing. Prior to this, Rosie worked on a range of projects prototyping future audience experiences and novel production tools for BBC R&D. Her background is in Computer Science (MSc), Physics (BSc) and Philosophy. In her spare time, Rosie runs the community group 'Manchester Futurists' which aims to explore the impact of technological acceleration on future society.

Abstract:

As we move towards ubiquotus video capture, content-producers face both challenges and opportunities. From UHD cameras, to smartphones, to CGI, we can use increasingly diverse sources of media to tell stories, but we must deal with disparate quality. We can give audiences interactive narrative experiences, but we must develop intuitive interfaces to author them. As our sources and archive vastly increase, we must find efficient ways to navigate and filter content of interest.

This talk will give an overview of some of the main projects BBC R&D are working on to overcome these challenges. We will explore areas such as at 'object-based media', dynamic video in the browser, and the EU project COGNITUS. We will then focus on a project that aims to create an artificially intelligent production system for live events. As part of this, we will examine approaches for augmenting human creativity and teaching machines the aesthetic principles of filmmaking, such as machine learning algorithms and neural networks. We will discuss progress so far as well as open research questions.