Japanese scientists said that they used artificial intelligence technology to successfully create spiritual images of the world's first objects and landscapes based on human brain activities.
The team of scientists from the National Institute of Quantum Science and Technology of Japan and the University of Osaka University can make rough images of leopard. Leopard has recognizable mouths, ears and spots patterns, and planes with red lights on the wings.
This technology is called "brain decoding", which can be visually visualized according to brain activity, and can be applied to the medical and welfare field.
The results of this research have recently been published on the online version of the International Scientific Magazine "Neural Network".
Earlier studies have shown that the images seen by human participants can be rebuilt through functional magnetic resonance imaging (FMRI), although they are limited to specific areas, such as letters.
Based on the previous method, the team has developed a technology that can quantify brain activity and allows generating artificial intelligence to draw images, combined with prediction technology to reproduce complex objects.
During the research, researchers showed the participants of 1,200 images of objects and landscapes, and used functional magnetic resonance imaging to analyze and quantify the relationship between their brain signals and images.The same image is entered into an artificial intelligence to understand the corresponding relationship between them and brain activity.