Enthusiast Creates Neural Network-Based Program that Picks Sound For Art Pieces
Sputnik/ Peter AllenTech21:47 14.02.2018(updated 21:49 14.02.2018) Get short URL
An author used a long time created neural network, to detect objects depicted on the pictures and a free database to create sound for the masterpieces.
Nao Tokui, an enthusiast from Japan decided that famous art masterpieces would only win, if they are followed by a proper background sounds and thus he made his own program, that using a neural network technology, analyzes the pictures and plays sounds that are appropriate. For example, if sea is depicted, it plays sounds of waves, if there are people on it — indistinctive talks recordings and so on.
READ MORE: AI Generates New Harry Potter Chapter and Netizens Can’t Get Enough of It
It didn’t require much — he used a neural network Soundnet, which by analyzing thousands of videos learned to detect certain objects on the videos, using visuals and sounds. He also used free online picture gallery and soundbank.