In a new milestone, CEBRA was able to map the neural activity of the mouse while watching the video, managing to predict and reconstruct the images with a 95% success rate.
Researchers at the École Polytechnique Fédérale de Lausanne (EPFL) used artificial intelligence to decipher what a group of mice were seeing thanks to their brain waves.
The case is another leap forward in the field. Just a few days ago, another artificial intelligence managed to decode in humans what they were thinking and convert it into text .
The case
The scientists developed an artificial intelligence tool called CEBRA that could interpret signals from the mouse brain .
This machine learning algorithm was trained to determine and parent specific frames based on brain waves.
The rodents were shown a 1960s movie, in a black and white scene where a man runs towards a car.
According to the study published in Nature , CEBRA measured and recorded mouse brain activity using electrode probes inserted into the visual cortex region of their brains.
They also genetically modified the mice so that their neurons will glow when they make synapses and can be easier to research.
The result
The artificial intelligence reconstructed what the mouse was observing , and the footage was very similar to what the animal viewed. This revealed that the AI could predict the correct frame within a second 95% of the time.
“Training the AI on data from multiple animals actually makes the predictions stronger, so you don’t need to train the AI on data from specific individuals to make it work for them,” says study author Mackenzie Mathis.
In addition, he affirms that this could help people with visual disabilities since visual sensations could be generated through patterns of brain activity.