Realeyes’ facial coding technology currently measures the seven basic emotions (e.g. happy, surprise, scared etc.) as well as engagement. However, the grant will help develop the ability of standard computer webcams to automatically detect whether a person likes or dislikes what they’re seeing.

The company said ‘likeability’ can not currently be measured by automated emotion measurement technologies.

The project, ‘SEWA: Automatic Sentiment Analysis in the Wild’, builds on the largest worldwide database of how ‘likeability’ is linked to emotional reactions (i.e. does happiness or surprise mean an ad is liked?). Realeyes has recorded the emotional responses of more than 60,000 people in conjunction with answers to a series of ‘likeability’ questions from sharing content to recommendations. The grant will help develop automated technology that will quantify this link between likeability and emotions.

Professor Maja Pantic from Imperial College will lead the research, with guidance from an industry board of global leaders in their respective fields – AOL (video advertising), Ipsos (consumer research), Kaplan (educational games) and Skype (video conferencing).