NEWYou can now listen to Fox News articles!
Researchers at Vanderbilt University have developed artificial intelligence technology to potentially help basketball players improve their game on the court.
Jules White, Associate Dean for Strategic Learning Programs and Associate Professor of Computer Science and Computer Engineering, and Carlos Olea, Ph.D. student in the computer science department, developed an artificial intelligence software called temporal relational network to help determine the context and mechanics behind every shot taken by a player.
“I’m really excited about the potential of AI to help home hobbyists learn and improve,” White told Fox News Digital. “I think a lot of people have limited time and money to access coaching, so I think it expands access to the information you need to improve.”
CLICK HERE FOR MORE SPORTS COVERAGE AT FOXNEWS.COM
The idea was born after a former student connected White and Olea with Basketball NOAH, a company that uses state-of-the-art cameras and software to watch practices and collect shooting stats for NBA and NCAA players. Using facial recognition and computer vision, NOAH generates detailed statistics on how and where someone shoots, the trajectory of the ball and where they missed their shots from.
NOAH provided more than 50,000 hours of video footage, which the researchers used to categorize five different types of shots: free throw, catch and shoot, out of dribble, fake misstep, and backing up. Thanks to AI, they were able to optimize the recognition of the type of shot, achieving an accuracy of 96.8% on 1,500 new shots.
White said the researchers were interested in knowing the stats, but also the context behind the stats, like if the player was just standing there shooting over and over, versus if they were in the middle of the game or s he had just received a pass. .
“We connected and discussed a research project where we would look at developing deep learning models to help them identify what was going on,” White said. “Basically, the AI is trying to learn and understand what movements are related to different types of shots. We basically teach him what things look like on video.
White said doing that is a human job in basketball because people in the sport have to watch those thousands of videos and individually label each one based on the type of shot.
He said Olea’s role was to use all of these examples to teach the AI how to identify the type of shot, with the seeker code basically telling the AI what it did wrong and where it went. made mistakes. White called it an interactive process where the AI receives examples and tries to predict what the correct answer is for those examples. After that, the program goes back and sees what it did wrong, and it updates its understanding of those moves.
“The AI we designed can be seen as a very naive person trying to learn something, in this case how to identify these different types of basketball shots,” Olea said.
CARDINALS DIRECTOR OLIVER MARMOL LAUNCHES ED TECH COMPANY TO MENTOR Aspiring ATHLETES
White said you can probably get the same kind of information by following a player with a clipboard, although he noted it would be “hugely expensive to do so”.
“For some players you can justify that expense, but for everyone else it can automate that process of gathering that information that the player and the coach can use to make decisions about how they can train.” , did he declare.
Olea said the simple categorization of shots is only a small part of basketball, as other factors can impact the shot, such as the quality of the defender or the speed of the player.
The researchers said they started with the basics, but hoped to evolve — with more research — into other parts, possibly assigning attributes to different actors. They said the partnership with NOAH was very helpful in understanding the stats and context behind each shot.
“We started with those first five things (the five types of plans), but to be effective from a research perspective, you kind of need a partner like that who really understands what matters and what are the stats you want to know and what are the contextual things,” White said. “So in the future you would want exactly things like that.”
Over time, White said he hopes the sports world will see more of this type of research. He thinks companies, like NOAH, could eventually take this kind of approach and apply it to real-world products and business systems you can use.
“So I bet over time the kind of advanced stuff we’re talking about here could be done on your phone,” he added, saying it could be an app installed on your phone, which people can use at home or at the gym. “I think the long-term vision as AI gets more sophisticated is that it would head in a direction like that.”
“That’s kind of where it gets exciting to just take something very raw that people are collecting now, like videos of themselves doing something…and give you that extra bit of information. to help you improve,” he added. “So that’s where I would love to have something like that for all things sports related.”
Olea, who grew up spending hours playing baseball, sometimes alone, said the technology would have served him well in his youth.
CLICK HERE TO GET THE FOX NEWS APP
“Having that kind of feedback would be extremely helpful in improving my own game and just understanding what I need to do to improve as a player,” he added.
White and Olea discussed shot analysis in an award-winning conference paper titled “Deep Learning Action Recognition Analysis for Basketball Shot Type Identification.”