IBM taps emotions for Wimbledon highlights
John McEnroe, known for his on-court outbursts, would be in his element at next month's Wimbledon tennis tournament, where every fist-pump of elation and cry of angst will be analysed by artificial intelligence (AI) to compile video highlights.
The All England Lawn Tennis Club and its technology partner IBM said the Watson AI platform had been taught to recognise players' emotions, which it would combine with an analysis of crowd noise, players' movements and match data.
IBM's Sam Seddon said the company was using machine learning to try to find those seconds when players had a heightened sense of emotion after an exciting shot or rally.
"If you've got the visual element from the player, and you know it's a tight pressure point in the match, then those are the points you are going to really target in on in the highlights package," he said in an interview.
He said the AI platform also analysed noise from the thousands of spectators who attend the annual tournament in South-West London.
"We are listening for the volume of the crowd; if you have got a fantastic rally, noise levels will go up and down during that point," he said.
"When that is happening, we flag it as particularly interesting."
Wimbledon was an early adopter of AI, Seddon said, and had widened its uses since it was first deployed in 2015.
This year, for example, IBM's Watson AI technology will power a Wimbledon Chabot service, delivered within Facebook Messenger, he said. It will allow fans access to tailored information on scores, news and players.