AI Brain Scans recreate images that people saw

An artificial intelligence system created by the University of California, Berkeley researchers can see what individuals are thinking by examining their brain activity. According to the article of Sarah Kuta (Daily Correspondent for Smithsonian Magazine) from March 2023, the technology creates real-time visuals of what the user is seeing by using data from fMRI scans to track activity in the visual cortex as the user views a sequence of images.

While volunteers were inside an fMRI scanner, the researchers presented them with a collection of 100 pictures to train the system. Faces, animals, and commonplace items were all shown in the photos. The AI system observed the participants' brain activity as they looked at the photos and eventually came to link particular patterns of activity with particular images.

The participants were shown a fresh set of images while they were inside the fMRI scanner to evaluate the AI system after it had been trained. The AI system created visuals that closely mirrored the images the participants were viewing after real-time analysis of the participants' brain activity. The visuals the AI system produced in several instances were quite similar to the ones the participants were seeing.

The researchers think that a variety of uses for their AI system are possible, ranging from assisting those who are unable to communicate verbally to advancing our knowledge of how the brain interprets visual information. The technique may potentially be utilized to create brain-computer interfaces that work better and let users think to control computers or other gadgets.

The researchers do admit that the system still has its limits, though. For instance, the AI system can now only replicate basic graphics; it is unknown how it will fare with more complicated ones. The method is said to rely on fMRI scans, which may be costly and time-consuming to conduct, according to the researchers.

Overall, the researchers think that their AI system is a fascinating advancement in our understanding of how the human brain works and how we perceive our environment.