Think about those moments when you’re watching your favorite sports team, and the referee makes a bad call that gets you all worked up. Well, Ayush Pai, a big NBA fan, decided to do something about it.
Ayush noticed that sometimes the refs miss important fouls, and the game’s outcome depends too much on their decisions rather than the players’ skills. So, using his self-taught knowledge of computer vision Artificial Intelligence, he created the first version of a tool for referees, calling it version 1.0. He did this right after finishing high school in the summer of 2022.
Ayush got the idea from systems like soccer’s VAR and tennis’s Hawk-Eye, which help referees make better decisions. He thought, “Why not do something like that for basketball?”
He realized that basketball is super fast-paced, especially in the NBA, and referees don’t always see everything that happens on the court. So, he developed a system that uses multiple cameras to process and detect trends in real-time, making it easier for court prosecutors to make accurate calls.
In simple words, Ayush wanted to provide a high-tech assistant referee to ensure that no important games were missed.
So, in the first version of Pai’s system, it could tell the color of the basketball Artificial Intelligence and count steps using a pedometer. But there were problems like bad lighting and a clunky step counter that could have been better. Then, in June, the Denver Nuggets won the NBA championship, but there were some referee mistakes in one of the games. It got Pai thinking, and he went back to work.
By the end of the month, he came up with version 2.0 of his Artificial Intelligence referee system. To make it better, he used more than 3,000 pictures of basketballs to teach the system how to follow the ball’s movements. He also used pose estimation to keep an eye on where players’ ankles and wrists are, helping the system catch things like double dribbles and traveling violations.
Now, Pai is already working on version 3.0. This new version will be able to handle multiple players, cameras, and even spot shooting fouls and reach-ins. And get this – he’s also creating a phone app that he hopes will bring fair officiating to youth leagues and casual pickup games all around the world. Pai’s on a mission to make sure everyone gets a fair shot at enjoying their game!
The AI trend is still going strong, and Ayush Pai thinks his creation can help more people understand and get into AI. He believes that everyone should know a bit about AI and how it can be useful in different areas, like ChatGPT, Tesla’s self-driving cars, and even in sports.
There’s this ongoing argument about AI possibly taking over jobs from people. But Ayush sees his system as something that can work together with humans instead of replacing them. He thinks it’s more about humans and AI teaming up, making things better, especially in the world of sports officiating.
“You definitely don’t want Artificial Intelligence to seem like it’s taking jobs away from people. It’s way better to think of it as a teamwork thing between AI and the folks doing the job. Like, in basketball Artificial Intelligence, there’s a lot of physical stuff that AI just can’t handle,” he explained.
Ayush from Florida is switching from the University of Florida to Georgia Tech this fall. He’s excited to start his second year at the new place.
“I moved to this new place mainly because they’ve got awesome opportunities in AI and computer science,” he shared. “Being part of the College of Computing’s Threads program will be a big boost for my projects. And I’m super excited to check out the entrepreneurial stuff here – they’ve got cool programs that help you come up with new ideas.”
Pai isn’t just into AI for basketball; he’s done other cool things, too. Like, he made systems that try to stop people from texting and driving, and he’s got stuff to help students stay focused while studying.
And check this out: Georgia Tech is becoming a big deal in the AI world. They just launched the AI Hub on campus, which is all about teaching and researching AI and making it useful in the real world while keeping things responsible. They’re making moves!