Quantum Neural Networks: Uncovering the Secrets of Human Perception Using Eye Errors

Eviralnews,; Optical errors, quantum mechanics, and neural networks may at first seem like completely unrelated topics. However, new research suggests a phenomenon called “quantum tunneling” for designing a neural network that can “see” eye errors in the same way as a human.

The new neural network designed by the researchers has used the two famous concepts of “Necker's cube” and “Robin's pot” in simulating human perception, and it has actually performed better than some normal and much larger neural networks used in computer vision. This model may be the answer to the question of whether artificial intelligence systems can really achieve something like human cognition.

The importance of visual errors

Optical illusions trick our brains into seeing things that may not be real. We still don't fully understand how optical illusions work, but studying them can help us understand how the brain works and how they might work in things like dementia or on long flights.

Researchers using artificial intelligence to mimic and study human vision have found that eye errors are problematic. While computer vision systems can recognize complex objects such as artistic paintings, they often cannot understand eye errors. Recent models appear to detect at least some of the errors, but these results require further investigation.

How does a quantum neural network work?

When the human brain processes information, it decides which data is useful and which is worthless. The new neural network mimics the functioning of the brain using many layers of artificial neurons that enable it to store and classify data as useful or not.

Neurons are activated by signals from neighboring neurons. Imagine that each neuron has to climb the wall of its neighboring neuron in order to light up, and in the same way other neurons have to pass the signal to neighboring neurons to finally reach the activation point on the other side.

In quantum mechanics, tiny objects like electrons can sometimes pass through seemingly impenetrable barriers through an effect called “quantum tunneling.” In the new neural network, quantum tunneling allows neurons to sometimes jump directly over a brick wall to an activation point, even firing when they “shouldn't”.

The importance of quantum tunneling

The discovery of quantum tunneling in the early decades of the 20th century allowed scientists to explain natural phenomena such as radioactive decay that seemed impossible according to classical physics.
Now, in the 21st century, scientists are facing a similar problem: existing theories explaining human perception, behavior, and decision-making are no longer valid.

Research has shown that the tools of quantum mechanics may help explain human behavior and decision-making.

While some believe that quantum effects play an important role in our brains, even if this is not the case, the laws of quantum mechanics may be useful for modeling human thinking. For example, quantum computing algorithms are more efficient than classical algorithms for many tasks.

With this in mind, the researchers tried to understand what would happen if they injected quantum effects into the functioning of a neural network.

How the quantum tunnel network works

When we see an optical illusion with two possible interpretations (such as a Necker cube and a Rubin vase), researchers believe that we temporarily hold both interpretations at the same time, until the brain decides which image to see.

This situation is similar to the thought experiment of quantum mechanics “Schrödinger's cat”. This famous scenario describes a cat in a box whose life depends on the decay of a quantum particle. According to quantum mechanics, as long as we observe the particle, it can be in two different states at the same time, and thus the cat can be alive and dead at the same time.

The researchers trained their quantum tunneling neural network to distinguish between a Necker cube and a Rubin vase. This model, when faced with an input illusion, produced an output of one of two interpretations. Over time, the interpretation he chose fluctuated. Traditional neural networks behave this way, but in addition, the new neural network produced ambiguous results between two given outputs: just like our brain, which can hold both interpretations before making a choice.

In the age of fake news, understanding how our brain processes illusions and constructs models of reality is of great importance, and accordingly, researchers are trying to see how quantum effects might help us understand and transform social behavior. Comments on social networks help.

This report has been translated into Farsi from the Conversation website.