Scientists use AI to read mouse brains and reconstruct scenes seen by mice
May 9, 2023
0
researchers from Federal Polytechnic School of Lausanne (EPFL) developed a machine learning algorithm that deciphers the brain signals of mice and reproduces images of what they see. Though
researchers from Federal Polytechnic School of Lausanne (EPFL) developed a machine learning algorithm that deciphers the brain signals of mice and reproduces images of what they see. Though secret mind reading is still a long way off, it is an exciting development that could improve our understanding of the human and animal brain in terms of how it responds to visual and other stimuli.
05/09/2023 at 14:00
News
Spotify removes thousands of AI-generated songs from Boomy
Streaming Platform Takes Action Against Artificial Streaming in Music Industry
Since 2011, scientists at the University of California have already created rudimentary videos of people’s brain activity inside MRI machines. In 2013, researchers at Radboud University in the Netherlands determined the letter a subject looked at based on brain activity.
In the EPFL study, mice watched a 1960s black-and-white clip of a man running to a car and opening the trunk. While watching the clip, the scientists recorded the brain activity of the mice using two approaches: probe electrodes inserted into the visual cortex of their brains, and optical probes for genetically modified mice whose neurons glow green when they transmit information. This data was used to train a new machine learning algorithm called CEBRA.
How SEBRA acted
With reference to brain signals received from another mouse that first watched a black and white clip, the CEBRA algorithm was able to correctly identify certain frames that the mouse saw while browsing. CEBRA was also able to create matching frames that were near perfect, but with some AI-generated image distortion.
Does this mean that a mouse with a brain probe can be used as a covert spying tool, and remote reading of its brain will be used to decipher everything it sees? Probably no.
This search included a very specific clip that the machine learning algorithm also knew about. CEBRA also takes into account only about 1% of neurons in the mouse brain, so there is room for improvement in its accuracy and capabilities. In addition, the research is not limited to deciphering what the brain sees, but can also be used to “predict hand movements in primates” and “recover the position of mice during free running in an arena.”
This is potentially a much more accurate way to look at the brain and understand how all neural activity correlates with information processing.
Mundo Conectado Deal Center: selection of discounts and lowest prices Best deals on electronics, cell phones, TVs, soundbars, drones and more
Donald Salinas is an experienced automobile journalist and writer for Div Bracket. He brings his readers the latest news and developments from the world of automobiles, offering a unique and knowledgeable perspective on the latest trends and innovations in the automotive industry.