[ad_1]
Apple has quietly acquired WaveOne, a Mountain View-based startup that develops AI algorithms for compressing video.
Apple would not confirm the sale when asked for comment. But WaveOne’s website was shut down around January, and several former employees, including one of WaveOne’s co-founders, now work in Apple’s various machine learning teams.
In a LinkedIn post published a month ago, WaveOne’s former head of sales and business development, Bob Stankosh, announced the sale.
“After nearly two years at WaveOne, we completed the sale of the company last week,” Stankosh wrote. “We started our journey with WaveOne realizing that machine learning and deep learning could change the world of video technology. Apple saw this potential and jumped at the chance to add it to their technology portfolio.”
WaveOne in 2010 Founded in 2016 by Lubomir Bordev and Oren Rippel, it aims to take decades-old video codecs and make them AI-powered. Prior to his tenure, Bordev was a founding member of Meta AI’s research department, and he and Rippel worked on Meta’s computer vision team responsible for content moderation, visual search, and feed ranking at Facebook.
Regarding standard algorithms for video compression and decompression, the compression is done by the content provider (e.g. YouTube’s servers) and the end users’ machines handle the decompression. It’s an effective approach, but new codecs require new hardware specially designed to speed up compression or decompression, making updates slower.
Image Credits: WaveOne
WaveOne’s main innovation was a “content-aware” video compression and decompression algorithm that runs on AI accelerators built into many phones and increasingly on PCs. Using AI-powered scene and object detection, the startup’s technology can “understand” a frame of video — allowing it to prioritize faces at the expense of other elements in a scene to save bandwidth, for example.
WaveOne also claims that its video compression technology is robust to sudden connection interruptions. That is, it can make a “best guess” based on the available bits, so when bandwidth is suddenly limited, the video won’t freeze – it will show a small detail for the duration.
WaveOne’s approach – which was hardware-agnostic – could cut the size of video files in half, with better returns on more complex scenes.
Investors saw the potential, apparently. Prior to the Apple acquisition, WaveOne attracted $9 million from backers including Khosla Ventures, Vela Partners, Incubate Fund, Omega Venture Partners and Blue Ivy.
So what can Apple expect from AI-powered video codecs? Well, the obvious answer is more efficient streaming. Even modest improvements in video compression can save bandwidth costs or allow services like Apple TV+ to offer higher resolution and framerates depending on the content being streamed.
YouTube is doing just that. Last year, the video-sharing service reduced the amount of data it needed to distribute to users by 4% by adapting Alphabet’s DeepMind’s machine learning algorithm to the problem of compressing YouTube videos to play board games.
Perhaps we’ll soon see similar innovations from Apple-owned WaveOne Group.
[ad_2]
Source link
