For most sports organizations, highlight videos are the bread and butter of digital content. With so many hours of game film, however, it can be difficult keep up with the demand for highlights.
That’s the challenge that Reely aims to solve. The platform automates visual content in sports and esports using in-house live feeds from teams and streaming services.
“It’s all about computer vision and artificial intelligence,” said Brent Marcus, Reely’s chief technology officer. “So, how the computer watches a video and breaks it into parts to be analyzed – that’s the combination.”
Reely’s current partners include Major League Lacrosse, the University of Kansas and the St. Louis Blues, as well as a variety of streaming providers. It has the ability to provide highlights for numerous sports, including soccer, lacrosse, football, basketball, baseball, and volleyball.
“There are other products out there offering automated sports events and highlights,” added Ian Stephens, co-founder and chief marketing officer of Reely. “What they do is produced by people in rooms, watching games, timestamping everything — it requires a human component. We’re building a computer vision platform that can recognize a touchdown without the human component. We’re using computer vision to watch events and understand them.”
Here’s how it works: the video feed connects with Reely’s system, which automatically recognizes big moments and tags them. The individual highlights are stored on Reely’s cloud-based software and can be searched and sorted. The highlights can then be packaged for distribution on social media or other digital platforms. The system works in real time, so the highlights are ready to go just seconds after an important play occurs.
Marcus explained that the system takes into account a combination of factors to give plays a “hype score.” Those factors include the tone and volume of an announcer’s voice, changes on the stream’s scoreboard, replays, and camera actions – all signs that a big play has occurred.
In an increasingly visual world, Stephens said that Reely’s services may be more necessary than ever.
“Right now, we’re creating more videos each day than ever before, but humans are super inefficient,” Stephens said. “We’re combing through video data, but there’s lost content because we’ll never be able to make it through it all.”
Reely addresses that problem by eliminating the need for human labor. It doesn’t just capture the action on the field or court, either – Reely is especially useful in the esports industry and is used for games like “Fortnite.”
“Esports gamers are sitting down for eight hours of content a day individually,” Stephens said. “They want to know, ‘What did I do in this game?’ They want all that data, just like athletes. There’s a much greater need because it’s literally impossible for them to edit all this content.”
Right now, Reely’s main focus is obtaining more users so the program can continue to evolve and learn. With more videos coming in, the company can create data sets for training modules, so the system learns to look for more specific moments in an event.
Although the company is still relatively young, Stephens and the rest of the Reely team understand the impact of maximizing the efficiency of the digital world using AI.
“It’s the inevitable future in the video space,” Stephens said.