New research demonstrates that a multitude of visual illusions stem from the limitations inherent in the functioning of our eyes and visual neurons, rather than being influenced by more intricate psychological mechanisms.
In a study published in the journal PLOS Computational Biology, researchers specifically investigated illusions wherein the color or pattern perception of an object is influenced by its surrounding environment. The paper titled "A model of color appearance based on efficient coding of natural images" delves into this examination.
For quite some time, scientists and philosophers have engaged in a longstanding discourse regarding the origins of these illusions. One school of thought suggests that they arise from neural processing occurring within the eye and the brain's low-level visual centers, while another perspective posits that higher-level mental processes like context and prior knowledge play a role.
Dr. Jolyon Troscianko, a researcher from the University of Exeter, collaborated on a recent study introducing a model that proposes that the illusions mentioned earlier can be explained by straightforward restrictions in neural responses, rather than more complex psychological mechanisms.
"Our eyes send messages to the brain by making neurons fire faster or slower," said Dr. Troscianko, from the Center for Ecology and Conservation on Exeter's Penryn Campus in Cornwall. "However, there's a limit to how quickly they can fire, and previous research hasn't considered how the limit might affect the ways we see color."
The model integrates the concept of "limited bandwidth" in neural processing with insights into human perception of patterns across various scales. Additionally, it assumes that our visual system is most efficient when processing natural scenes.
The two gray bars in the middle of this figure are the same gray, but the one on the left (surrounded by more black bars) appears darker. This is the opposite of the simultaneous contrast example above, because darker surrounds now make the target look darker. Credit: Jolyon Troscianko
Researchers from the Universities of Exeter and Sussex collaborated on the development of a model primarily aimed at predicting how animals perceive colors. However, during the course of their study, they made an interesting discovery—the model also accurately predicted numerous visual illusions experienced by humans.
Dr. Troscianko remarked that these findings challenge long-standing assumptions regarding the mechanisms behind visual illusions. Additionally, the research sheds light on the popularity of high-definition televisions.
He explained that modern high dynamic range televisions generate intensely bright white areas that surpass the darkness of black by over 10,000 times, approaching the contrast levels observed in natural scenes.
The capability of our eyes and brains to handle such extreme contrast presents a puzzle, as studies indicate that the maximum contrast humans can perceive at a single spatial scale is approximately 200:1. Intriguingly, the neurons responsible for transmitting visual information from our eyes to our brains can only handle contrasts of around 10:1.
The model developed in the study illustrates how neurons with such limited contrast bandwidth can combine their signals, enabling us to perceive these immense contrasts. However, this information becomes "compressed," leading to the manifestation of visual illusions.
Furthermore, the model demonstrates how our neurons have evolved to utilize their capacity to the fullest extent. For instance, certain neurons are highly sensitive to subtle variations in gray levels at medium-sized scales but can be easily overwhelmed by high contrasts. Conversely, neurons that encode for contrasts at larger or smaller scales exhibit lower sensitivity but can operate over a much wider range of contrasts, resulting in distinct black-and-white distinctions.
Ultimately, the research illustrates how a system with severely limited neural bandwidth and sensitivity can perceive contrasts exceeding 10,000:1.
A model of colour appearance based on efficient coding of natural images, PLoS Computational Biology (2023). DOI: 10.1371/journal.pcbi.1011117