LAS VEGAS — Manually tagging your videos with all the possible people, things and phrases that appear within would be a time-consuming task for anyone, much less for content owners operating at scale.
So Microsoft has brought its Cognitive Services machine learning suite to bear on video. Hosting a booth at NAB Show, Microsoft Azure Media Services principal program manager Martin Wahl explains to Beet.TV it means cloud technology used to power features like:
- Facial analysis
- Voice recognition
- Spoken word recognition
- Written word identification
- On-screen word understanding
- In-motion object detection
These are some advanced tools. Azure was already powering a computer vision API to understand static images and a face API to detect faces in photos, as well as various other APIs like speech and translation.
Now it is joining them with a video API, currently in preview, claiming to produce stable video output, detect motion, create intelligent thumbnails and detect and tracks faces.
Not only that, but Wahl says the tech can work in real-time – so that content owners can and advertisers can make real-time decisions, translate to another language or develop interactive apps around videos, as they play out.
Ooyala has integrated with the technology for an intelligent video suite it announced ahead of NAB Show.
This video is part of Beet.TV’s coverage of the 2017 NAB Show in Las Vegas. The series is sponsored by Ooyala. For more coverage of NAB, please visit this page.