Artificial ‘brain’ may reveal the mystery behind optical illusions

26 Feb 2021

Image: © Tartila/Stock.adobe.com

A new computer system delves into how the human brain processes moving pictures and why we can’t always believe our own eyes.

While the human eye can judge moving objects with an accuracy that allows us to catch a ball, many aspects of this visual motion perception are still not understood.

Using decades of research about how the human eye processes moving images, researchers have trained an artificial ‘brain’ to match this ability, revealing new insights into optical illusions.

In a study, published in the Journal of Vision, the team from University of Cambridge designed an artificial system, MotionNet, to closely match the motion-processing structures inside the human brain. This allowed the team to examine certain features of human visual processing that cannot be measured directly in the human brain.

One of the insights researchers were able to delve into was the mystery behind optical illusions and how the human brain can be fooled by them.

For example, if there’s a black spot on the left of a screen, which fades while a black spot appears on the right, humans will believe they can ‘see’ the spot moving from left to right. This is called ‘phi’ motion.

However, if the spot that appears on the right is white on a dark background, humans will ‘see’ the spot moving from right to left, in what is known as ‘reverse-phi’ motion.

‘A completely new prediction’

The researchers reproduced reverse-phi motion in the MotionNet system and found that it made the same mistakes in perception as a human brain, allowing the team to examine why it happens.

They found that neurons are ‘tuned’ to the direction of movement and, for the artificial system, reverse-phi was triggering neurons tuned to the direction opposite to the actual movement.

The system also revealed that the speed of the reverse-phi motion is affected by how far apart the dots are, but not in the way that might be expected. It found that dots that ‘moved’ at a constant speed appear to move faster if spaced a short distance apart, and more slowly if spaced further apart.

Dr Reuben Rideaux, a researcher in the University of Cambridge and first author of the study, said the new model generated “a completely new prediction” about how humans experience reverse-phi.

“It’s very hard to directly measure what’s going on inside the human brain when we perceive motion,” he said. “Even our best medical technology can’t show us the entire system at work. With MotionNet we have complete access.”

Researchers are confident that MotionNet can be used to fill in many more gaps in the current understanding of how the human brain perceives motion.

While the system’s predictions will need to be validated in biological experiments, the team said knowing which part of the brain to focus on will save a lot of time.

Jenny Darmody is the editor of Silicon Republic

editorial@siliconrepublic.com