summary: Using a combination of machine learning and neuroimaging data, researchers have uncovered the neural underpinnings of aesthetic perception.
sauce: caltech
It has been said that the taste is inexplicable. But what if taste could actually be explained, and what if that explanation was done by a neural network in your brain?
In a new paper published in Nature Communicationsa team of researchers at the Caltech show how they used a combination of machine learning and brain-scanning equipment to uncover the neural underpinnings of human aesthetic preferences.
The work was done in the lab of John O’Doherty, the Fletcher Jones professor of decision-making neuroscience at the California Institute of Technology, and builds on research published by that lab in 2021. Data about the pictures volunteers liked and disliked. With enough training, computers can now correctly guess whether a person prefers Monet or Rothko, for example.
The act of liking or disliking a work of art seems innate and happens so instantly and seamlessly in our brains that it takes time to think about why or how it happens. Very few people would. Hundreds of years.
“When I see a picture, I immediately decide if I like it or not, but when you think about it, it’s very complicated because it’s very complicated to type,” says former Caltech and now Kiyohito Iigaya, lead author at Columbia University, said: .
“This is actually a very open question, and how the brain does it is not well understood, so I was wondering if we could use computational modeling techniques to understand it. “
In this method, volunteers rated paintings (1,000 paintings) over four days while their brains were scanned with a functional magnetic resonance imaging (fMRI) machine.
These brain scans and volunteer evaluations of the paintings are based on contrast, hue, dynamics, and concreteness (whether the painting is abstract or realistic).
The data the team collected showed that regions within the visual cortex, the part of the brain that processes visual input, are responsible for analyzing these properties. A region in the front of the brain known as the medial prefrontal cortex (mPFC) is responsible for assigning subjective values to them.
Basically, the brain breaks down a work of art into essential qualities and decides whether those qualities are pleasing or not. According to another study conducted by the O’Doherty Institute, this is pretty much how the brain decides whether you like a food or not. That study found that the brain analyzes foods according to their protein, fat, carbohydrate and vitamin content to determine if their quality is satisfactory.
“What they discovered is that the brain integrates these different nutritional characteristics to create an overall preference for a food,” says Iigaya. is the inspiration for.”
In the paper, the researchers say their findings suggest that this “value construction” system may be pervasive throughout the brain and may explain many kinds of preferences. said that
“I find it amazing that this very simple computational model can explain such a large difference in our tastes,” says Iigaya.
About this art and neuroscience research news
author: press office
sauce: caltech
contact: Press Office – Caltech
image: image is public domain
See also
Original research: open access.
“Neural Mechanisms Underlying Hierarchies of Perceived Aesthetic ValuesKiyohito Iigaya and others Nature Communications
overview
Neural Mechanisms Underlying Hierarchies of Perceived Aesthetic Values
Little is known about how the brain calculates the perceived aesthetic value of complex stimuli such as visual art.
Here, we use computational techniques in combination with functional neuroimaging to demonstrate that the aesthetic value of visual stimuli is determined through weighted integration of both low- and high-level stimulus features contained in early and late visual cortices. We provide evidence that hierarchically computed parietal and lateral prefrontal cortices.
Parietal and lateral prefrontal cortex feature representations may be utilized to generate the overall aesthetic value of the medial prefrontal cortex.
Such brain-wide computations are not only consistent with feature-based mechanisms for value construction, but also resemble computations performed by deep convolutional neural networks.
Our findings thus shed light on the existence of general neural computational mechanisms for generating value judgments rapidly and flexibly across a range of complex and novel stimuli and situations.
