Show Mobile Navigation

Thursday, March 09, 2017

, ,

Google's API Can Now Recognize Objects In Video

Oyetoke Tobi - Thursday, March 09, 2017

Google’s intelligence API, or application programming interface, will in a little while be able to recognize and differentiate objects in Google Cloud videos and will make those objects searchable.

Machine learning and AI has been Google’s core strength and focuses. And this has really reflected across its range of consumer products. The smart replies in Inbox, the ability of the Google Assistant to search for images from a specific keyword or phrase. Presently, Google wants to emphasis that its cloud platform is just as smart, and driven by Machine Learning tools that can be used by enterprise customers.

Google’s chief scientist for cloud and machine learning Dr Fei Fei Li, unveiled a new tool that could allow for computers to understand and decode a video, just how humans do; the new Video Intelligence API at the ongoing Google Cloud Next conference in San Francisco. Wednesday.

The API is still presently in private beta, so it’s not available to general users yet. However, there is a sign up for qualified users and developers. 

While training computers to recognize and understand images is something that Google has been good at, however, videos is another thing entirely. As a matter of fact, according to Dr Li, it is the ‘dark matter’ of the digital universe, but it now seems like Google has cracked how to decode some part of this. So basically, the Google’s new Video Intelligent tool, which is now in private beta, will able to recognize the exact part of a video that a user wants to find.

This API will allow for a deeper width of information about the contents of videos by recognizing and naming the contents with nouns and verbs. And each shot in a video will be searchable from within the videos that are stored in the Google Cloud, according to a release. 

So assuming you are looking for a video with trees, you’ll be able to search “trees” or “vegetation” or whatsoever search term you would use to find the shots of the item you’re looking for.

Google elaborated on it by uploading a number of videos that shows how the API works and lets users try out a demo of what it’ll be like when it’s completely released for general use. In its demo during the keynote address, Google showed how the tool could recognize and identify exact labels; when asked to find beach or baseball from a series of videos the tool was able to find exactly which clips had images of a beach/baseball and at what points.



With this, users would be able to search each shot, frame thanks to the tool, without searching it manually, in order to find the exact video footage


The API is very useful for organizing content on a diversity of levels. Google mentioned in their release that it can help companies sort and organize user generated content or to help companies create their own software for managing videos.

The video recognition feature is just one update to the Google API update that which also includes updates to DataLab and Cloud Jobs.



0 comments:

Post a Comment