Subscribe Now
Trending News

Blog Post

Show HN: I turned my face rec system into a video codec
News

Show HN: I turned my face rec system into a video codec 

Vertigo.ai Focus is a new from-the-ground-up patent pending AI & GPU video CODEC that enables hi-def face to face interactions over low-bandwidth links. It is production-ready, extremely light-weight, and able to run in real-time on devices as old as the iPhone 6s. Focus builds on the Vertigo Engine, which demonstrates the power and versatility of our platform.

The core video CODEC works in combination with multiple neural networks that perform head and face tracking and real-time face recognition, to either filter away unrelated bystanders, or to allow everyone in view of the camera to be present in the call on equal footing. The understanding of image contents afforded by the real-time AI allows for video bandwidth to be allocated to the eyes and mouth, areas of the face that research have shown to receive the most attention during face-to-face conversations. The images below demonstrate the dramatic difference this makes when comparing against Zoom on the same hardware:

Focus for IOS is now available on the App Store at https://apps.apple.com/app/vertigo-focus/id1540073203 , soon to be followed by a MacOS version (currently available on Testflight, please contact us for access). The core technology is also available on Linux and Windows, with support for Android in the works.

Please contact info@vertigo.ai for support, or jacob@vertigo.ai with any business queries.

Please see this list of 3rd party software used in Vertigo Focus.

Read More

Related posts

© Copyright 2022, All Rights Reserved