“In volleyball, we’re now using computer vision technology cameras to track not only athletes, but also the ball,” said Allen Jobrist, head of Omega Timing. “It’s a combination of camera technology and artificial intelligence.”
Omega Timing has 180 engineers in its research and development department, and according to Jobrist, the home’s internal development process began in 2012 with positioning systems and motion sensor systems. The goal was to reach a point where events for multiple sports in more than 500-plus events it works on each year could provide detailed live data on the performance of omega athletes. That data should take less than a tenth of a second to measure, process, and occur so that it matches what viewers are seeing on the screen.
Including beach volleyball, this means adopting this position and motion technology and training an AI to identify unforgettable shot types – from smash to block and spike and its various types, and the type of pass, as well as the flight path of the ball. Cut from the gyroscope sensor. These motion sensors allow the system to know the direction of movement of athletes, as well as the height of jumps, speeds, etc. Once processed, they are fed directly to the broadcast for use in all comments or on-screen graphics.
According to Jobrist, one of the most difficult lessons to learn for AI was being able to play the ball correctly when the cameras could no longer see it. “Sometimes, it is covered by the body part of an athlete. Sometimes it’s out of the TV frame, “he said.” So, the challenge was to track it down when you lost it. And got back, and fill [missing] Data and then continue automatically. That was the biggest problem.
This tracking of the ball is for ATI to determine what is happening during the game. “When you can track the ball, you know where it was and when it changed direction. The algorithm can detect the shot with a combination of sensors on athletes, “said Jobrist.” Whether it’s blocked or crashed. You know which team it was and which player it was.
Omega Timing claims that its beach volleyball system is 99 percent accurate, with sensors and multiple cameras running at 250 frames per second, thanks to Toby Bracken, professor of computer vision and image processing at the University of Durham, who is keen to see if it stands during the games – and perhaps, Cruc whether the system has been fooled by race and gender
“What has been done is reasonably impressive. And you’ll need a big data set to train AI for all the different moves, “says Brackon.” But one thing is for sure. How many times is it wrong with these different steps? How many times will it lose the ball track? Works uniformly for genders.Does the female team in the United States rely on 99 percent accuracy? And 99 percent accuracy in the Ghanian women’s team? “
Jobrist is confident, and explains that it may be easier to call Google or IBM for the necessary supplies of AI experts, but for Omega it was not an option. “The most important thing, whether it’s for a scoring sport or a timing sport, is that there can be no difference between us in the interpretation of performance and the final result.” So to protect the integrity of the results, we cannot rely on other agencies. We need to have the skills to explain the results and how the athletes got there. “
Jobrist has been heavily pressured about future scheduling and tracking upgrades, but says the 2024 Paris Games will be a key issue. “You will see a complete set of innovations. Of course it will be around timekeeping, scoring and of course speed sensors and positioning systems. And of course Los Angeles in 2026. We’ve got some really interesting projects out there that we’ve actually just barely started. “
More great wired stories