The neural network created images by “reading people’s minds”
[ad_1]
The neural network created images based on the thoughts of the participants in the experiment, who were inside the functional magnetic resonance imaging machine.
Scientists from Stanford University, the National University of Singapore and the Chinese University of Hong Kong invited volunteers to look at images that the program was supposed to reconstruct.
Not in all cases, the image looked similar to the one in front of the volunteer’s eyes: the owl looked like a monkey with large ears, the fish looked like a lizard, and the bat looked like a cross between an ordinary mouse and a kiwi bird. Nevertheless, in about 84 percent of cases, the program “guessed” the general theme, color scheme and texture of the image, biologists say.
[ad_2]
Source link