The following video is brought to you courtesy of the RT America YouTube Channel. Click the video below to watch it now.
Tech Times reports that Google researchers have created a ‘deep learning system’ that can single out voices by watching people’s faces when they’re talking. The researchers describe the new effort in a paper entitled, “Looking to Listen at the Cocktail Party.” In a blog post, they say that they can look at video of a bunch of people, select any one of their faces, and isolate what that one person is saying, removing the noise of everyone else talking, or sirens, or any other noises going on in the video. The Resident discusses.
Follow The Resident at http://www.twitter.com/TheResident
Find RT America in your area: http://rt.com/where-to-watch/
Or watch us online: http://rt.com/on-air/rt-america-air/
Like us on Facebook http://www.facebook.com/RTAmerica
Follow us on Twitter http://twitter.com/RT_America