This project investigates the feasibility with which a crowd of spectators at an event can create a compelling mobile outside broadcast for online viewers.
While there is a long-standing tradition of national broadcasters covering major events from the London Marathon to Glastonbury, it is not economically feasible to extend this to the very many similar but smaller-scale events across the country. Moreover, the traditional outside broadcasting model arguably fails to take into account an individual audience member’s shifting interests, quickly adapting to unfolding events, or providing significant coverage of a geographically large or diverse setting.
We propose a new service in which a crowd of spectators capture and broadcast live video using their own mobile phones, which is then streamed over the Internet for others to see. The role of an expensive professional production crew is taken on by a combination of social networking, recommender systems and location-based services, which select and recommend streams of likely interest to viewers, and also guide mobile camera operators as to where to go to get the best possible coverage. The net result is a crowd-sourced broadcast – a live YouTube with additional coordination and direction services, and the ability to select from potentially hundreds of viewpoints of a live event.
The project aims to address a number of research challenges. How can a crowd of spectators be engaged in creating a sufficient number of usable live video streams? How can a crowd of online viewers usefully act as a “director”? How can mobile phone sensors provide meaningful and real-time meta-data on the usefulness of mobile video stream? What are the privacy, ethical and innovation challenges implicit in deploying such a service?