Artificial intelligence, which is developing day by day, has started to be used in almost every field. A research team, which uses artificial intelligence, which is used in every conceivable field from health to transportation, from education to daily life, and now in the field of transportation, has developed an intelligence model based on human brain activity and has carried out an interesting study.
Researchers at Kyoto University conducted a study in the virtual reality world to measure the ability of subjects lost in a maze to predict their location from their brain activity and their ability to predict what they can see in the environment, and their confidence in their predictions. In a hypothetical scenario, the subject navigates a maze comparing his prediction to the scene he is observing, thereby confirming or updating the scene he has seen before.
The more people trust their guess, the easier it is to find the right path.
Imagine that the city you live in was destroyed by a disaster. Where you live can turn into a giant rubble maze, right? How could you find your home in an area with very few landmarks left after this disaster? A research team carried out a study that showed what the subjects could do in such a situation by transferring this situation to a virtual reality maze game.
Brain activity of subjects entering a virtual reality maze game was measured using functional magnetic resonance imaging and fMRI methods. The subjects, who had no information about their destination, used both their guesses and the memory of the map to choose their position in the maze and the correct path.
The lead author of the study, Risa Katayama, stated that the decoding accuracy of the scene prediction in this game, which is performed with an artificial intelligence model based on human brain activity, depends on the subject’s predictive ability and confidence in the prediction.
The results showed that when the confidence in the prediction was high, the subjects were able to clearly imagine the scene and predict it quickly. This study is expected to influence studies in the growing metaverse field.
8002
