Goalbeceepers struggle to guess which way a penyty penalty shoot
Javier Soriano / AFP by GETTY images
Deep learning models have been trained over 1000 penalty kicks in football matches can predict how mummy goongkeepers in life.
“Punishment kicks are some of the most decisive soccer moments, often determines the consequence of major tournaments,” as David Frier-Obregón At the University of Las Palmas de Gran Canaria, Spain. “However, real-time support for Goalheverers still Intuition-Base Base
So the Freire-obregón and his companions were charged with 1010 penalty kicks from real, television matches in Spain. In clips, 640 were considered analyzed by AI models, while the rest was thrown out of blow, it was short or obstructed.
Each clip is fed with 22 deep learning models, which should be aware of whether the penalty goes to the left, right or simply footage and the simple part of a player.
The best model model that can correctly identify whether the ball is toward the right, left or middle 52 percent of time with accuracy in time. If researchers have taken less than half-choice, model temporarily increases 64 per cent – almost 10 percent shepherds given to the same information.
Researchers wondered “how much subtle motion cues, before kicking the ball, the purpose can be revealed”, Freire-obregón says. He hopes that information can be useful for training goalseverepers, but using a AI prediction in a match situation can be more challenging.
“Our intention to explore the next if the possibilities expect the first for the punishment kicks, using the kicker’s action before the shot,” he said. “And if so, how early as such predictions will be made while still meaningful accuracy.”
Topics:
