[ad_1]
“In volleyball, we are using cameras with computer visual technologies not only for athletes, but also for tracking the ball,” says Alain Zobrist, head of Omega Timing. “So it’s a combination. We use camera technology and artificial intelligence to do that.”
Omega Timing’s R&D department has 180 engineers, and the development process began with home positioning systems and motion sensor systems, according to Zobristen, in 2012. The goal was to reach multiple points for multiple sports in 500-point sports. events that it deals with every year, Omega can provide accurate live data on athlete performance. This data would take less than a tenth of a second to be measured, processed, and transmitted at events so that the information matches what the viewer sees on the screen.
With beach volleyball, which began to take this positioning and movement technology and train an AI to know the different types of shots (from breaks to blocks to tips and their variations) and the types of passes, as well as the flight path of the ball, combine that data with gyroscope sensors in players’ clothing. These motion sensors allow the direction of movement of athletes, the height of jumps, speed, etc. The system is reported. Once processed, all of this is transmitted directly to the senders for use in comments or on-screen graphics.
According to Zobrist, one of the most difficult lessons the AI learned was to keep the ball in the game accurately when the cameras could no longer see it. “Sometimes it covers the body part of an athlete. Sometimes it’s outside the television frames, ”he says. “So the challenge was to keep track of when you lost the ball. The software can predict where the ball is going and then reappear, recalculate when you lost the object, and regain the gap, and fill it [missing] data and then continue automatically. That was one of the biggest problems. “
This ball tracking is key to determining what is going on in the AI game. “When you track the ball, you know where it was located and when the direction changed. And when combined with the athletes’ sensors, the algorithm will detect the shot, “says Zobrist.” It was a block or a smash. You know which team and which player it was. So this combination of the two technologies allows us to be accurate in measuring data. “
Omega Timing says the beach volleyball system is 99 percent full, thanks to sensors and multiple cameras that run at 250 frames per second. Toby Breckon, a professor of computer vision and image processing at the University of Durham, however, is interested in seeing if the Games will last, and especially if the system is deceived by racial and gender differences.
“What’s been done is pretty impressive. And you would need a big data set to train an AI in every different move, “says Breckon.” But one of the things is accuracy. How often does he go wrong with those different moves? How often does he lose track of the ball? Is there a 99 percent accuracy in the U.S. women’s team, for example? and 99 percent accuracy in the Ghanaian women’s team? “
Zobrist is confident, and explained that it might have been easier to call Google or IBM to provide the necessary AI specialization, which was not an option for Omega. “What matters most is whether it’s a scoring sport or a time sport, we can’t disagree between the explanation of performance and the end result,” he says. “So to protect the integrity of the result, we can’t trust another company. We need to be an expert in explaining the result and how athletes get there.”
As for future upgrades and follow-ups, Zobrist is tight, but says the 2024 Paris Games will be key. “You will see a new set of innovations. Of course, it will also be about time, score, and certainly motion sensors and positioning systems. And certainly Los Angeles in 2028 as well. We have some really interesting projects for that, because we just started. “
More great KABEKO stories
[ad_2]
Source link
