Apple acquires AI-powered video compression startup WaveOne

WaveOne, a California-based startup that has been developing algorithms for compressing videos, has confirmed that it has been purchased by Apple.

Founded in 2016, WaveOne sought to use machine learning for video compression. By employing artificial intelligence, WaveOne could create "content-aware" video compression and decompression that prioritized faces and text, and deprioritized background information like grass or trees.

It also claimed that its methods were hardware-agnostic. The company claims that the technology can reduce the size of video files by up to half while still retaining quality in complex scenes.

As noted by TechCrunch, Bob Stankosh, WaveOne's former head of sales and business development announced the deal on LinkedIn in February.

"After almost 2 years at WaveOne, last week we finalized the sale of the company to Apple. We started our journey at WaveOne, realizing that machine learning/deep learning video technology could potentially change the world," Stankosh wrote. "Apple saw this potential and took the opportunity to add it to their technology portfolio."

Apple has not confirmed the acquisition. The terms of the deal are currently unknown.

5 Comments

lkrupp 10521 comments · 19 Years

About 1 year ago

Do we really need more compression codecs? Aren't there enough to contend with already? How many do operating systems support them right now?

Xed 2896 comments · 4 Years

lkrupp said:

Do we really need more compression codecs? Aren't there enough to contend with already? How many do operating systems support them right now?

Of course we do. Getting more efficient codecs is a good thing. We had a lot of codecs when MPEG-2 was used for DVDs and I'm very glad we eventually got H.264 and H.265. I'm hoping that we see Versatile Video Coding (VVC, H.266, MPEG-I Part 3, ISO/IEC 23090-3), which will halve the size again moving from 64x64px macro blocks to 128x128.

setispock 2 comments · 6 Years

lkrupp said:

Do we really need more compression codecs? Aren't there enough to contend with already? How many do operating systems support them right now?

I don’t think they are inventing a new codec per se. Rather they are trying to use AI/ML to intelligently compress each scene. Important features like peoples faces would be lightly compressed whereas background scenes like the sky could have greater compression without a discernible loss in viewing fidelity.

dk49 284 comments · 9 Years

They might use this for upcoming VR headset.

fastasleep 6451 comments · 14 Years

lkrupp said:

Do we really need more compression codecs? Aren't there enough to contend with already? How many do operating systems support them right now?

Imagine being so dense that this is your takeaway from the article.

Share Your Thoughts on our Forums ->