Apple acquired a startup specializing in AI-based video compression earlier this yr, it has emerged, however it isn’t but clear how the corporate plans to make use of the know-how.

As is usually the case when Apple makes an acquisition, there was no official announcement from Cupertino, nor have the phrases of the deal been disclosed. On this case, nonetheless, a co-founder on the acquired agency, WaveOne, wrote in regards to the acquisition on LinkedIn, in a submit that TechCrunch found and reported on this week:

“Final week we finalized the sale of the corporate to Apple. We began our journey at WaveOne, realizing that machine studying and deep studying video know-how may probably change the world. Apple noticed this potential and took the chance so as to add it to their know-how portfolio.”

Apart from the LinkedIn announcement, which evidently didn’t make a lot of a splash when it was initially posted, TechCrunch has additionally noticed that a number of former WaveOne staff now work on machine studying at Apple. Regardless of an absence of official remark from Apple itself, the cat could be very a lot out of the bag.

In a profile of WaveOne again in 2020, TechCrunch defined how the corporate was utilizing machine-learning {hardware} like Apple’s Neural Engine to speed up and streamline video compression on cell gadgets.

What we don’t know for certain is how Apple sees the acquisition enjoying into its long-term technique, though there are apparent methods the corporate may benefit. AI is an space of rising significance inside the tech sector, however most firms are specializing in (or speaking about specializing in) artistic instruments alongside the strains of ChatGPT, and improved good assistants. WaveOne’s work, in the meantime, is on the machine studying facet and focuses on pragmatic options to particular issues. As TechCrunch explains, WaveOne’s flagship invention was a “content-aware” technique for analyzing video information and studying one of the simplest ways to compress them–prioritizing faces on the expense of boring background parts, for instance.

As helpful as it’s, that’s hardly prone to feed into developments to Siri, or to assist pave the best way to a ChatGPT competitor. However optimizing Apple’s TV+ streaming service by decreasing its information by only a few p.c may result in main value financial savings or enhancements to decision or framerates.