Creating the world's first mental images using artificial intelligence... How?

2023-12-20 2023-12-20T09:08:54Z

ArabiaWeather - Day after day, artificial intelligence technology amazes us with its awesome potential and its ability to fulfill human dreams that he has been aspiring to for years. Japanese scientists recently revealed an exceptional achievement, as they succeeded in creating the world’s first mental images of objects and landscapes using artificial intelligence technology, and the results were published. This research is in the journal "Neural Networks" .

A team of scientists from the National Institutes of Quantum Science and Technology , in cooperation with another national institute and Osaka University, led the process of creating approximate images of a tiger, with fine details such as the mouth and ears, and a recognizable spotted pattern, in addition to rendering objects such as airplanes with red lights. On its wings.

During the research phase, participants were shown 1,200 images , which included objects and landscapes, and the relationship between brain signals and images was analyzed using fMRI technology.

The images were fed into a generative artificial intelligence model to determine their compatibility with brain activity, which represents an innovative step towards exploring the amazing capabilities of modern technology in understanding and embodying the details of the human mind.

Brain decoding technology

“Brain decoding” technology offers the unique possibility of imaging cognitive contents based on brain activity. This innovation is referred to by its amazing name in the field of technology. This approach, developed according to "The Japan Times" website, allows it to be applied in several fields, including Medicine and social care.

This technology is also useful in developing communication devices, providing a deeper understanding of the mechanisms of hallucinations and dreams in the brain, according to the article. Based on previous research, the images that individuals see can be reconstructed by measuring brain activity using functional magnetic resonance imaging techniques, although these studies were previously limited to specific areas such as the letters of the alphabet.

By developing advanced methods, the team has succeeded in developing technology capable of measuring brain activity and enabling generative AI to draw accurate images combined with prediction techniques, enabling objects of high complexity to be recreated.