Accedo launches AI project with ITV and AWS

  •   
  •   
  •   
  •   

Accedo, the video experience specialist, has announced a project with British broadcaster, ITV  to use Artificial Intelligence to automatically generate cover artwork which is likely to appeal best to a given user segment. The project will be tested with a focus group of ITV users.

The project follows a prototype developed on top of the Amazon Web Services cloud platform to leverage the emotional triggers of viewers to deliver personalised cover art and in turn increase user engagement and video consumption. It has initially been tested on a number of video samples, generating thumbnails for different artwork strategies, such as focusing on main characters or the mood of the user, for example.

Working with ITV, Accedo will be able to validate the business case. The focus group with ITV’s customers will aim to evaluate in depth emotional reactions to a different set of pre-generated thumbnails. The observations will be the baseline for a prototype set of rules to preselect relevant thumbnails, such as key characters in the episode, key elements in the episode’s description, or key scenes, for example.

Lee Marshall, Head of Product, ITV Hub, commented: “Improving User Engagement is extremely important to us, helping to ensure that we are delivering relevant and interesting content to all of our users. This projects opens up huge potential to offer our viewers a much more personalised, and therefore engaging, experience.”

Michael Lantz, CEO, Accedo, added: “Video services are gradually putting more and more effort into personalizing services in numerous ways, from personalised ads to content recommendations. This technology has the potential to revolutionise that personalization. Collaborating with ITV gives us not only the content to test our prototype on, but also the reality check our innovation ideas need to make sure they address an existing challenge in a way which will have a positive impact on our customers’ business.”


  •   
  •   
  •   
  •   

You must be logged in to post a comment Login