Visual Summarisations for Computer-Assisted Live Color Casting and Direction in League of Legends.

Esports has become a juggernaut in mainstream entertainment. For broadcasting and production stakeholders to establish a foothold in this exploding market, live-streaming broadcasts of matches should become more meaningful for fans, and more attractive to newcomers, ensuring their retention. This work describes an end-to-end solution for enhanced live production and streaming of the popular esport, League of Legends . The solution is based on several interconnected dashboards, which are fed by an inference engine reasoning on events transpiring in the game. Its aim is to lend decision support to directorial and color commentary tasks undertaken by the production crew. Hence, it supports the broadcasting director in using data to create shots of rich visual storytelling, while generating automated running commentary for esports casters to adapt the broadcasting script on-the-fly. Deduced knowledge is linked to custom infographics, that can be used to overlay the stream with visual summarisations of the action. The tools comprising our solution integrate with professional-grade broadcasting software, leveraging established workflows, and enabling complex management of multiple scenes and graphics from a single point of access. Our solution’s usability has been evaluated by esports experts through a cognitive walkthrough, verifying the approach.