News Release

What numbers do you see? A new bistable perceptual phenomenon on symbolic numbers

Researchers from the National Institute for Physiological Sciences in Japan found that partial occlusion of digital numerals induces bistable interpretations of their semantic meanings. Moreover, long-time visual exposure to normal numerals biases percept

Peer-Reviewed Publication

National Institutes of Natural Sciences

The illustration of bistable interpretation of digital numerals and visual adaptation task paradigm

image: 

(A) The cover of the traffic countdown timer occludes part of the digital numerals. (B) The partial occlusion induces bistable interpretation of the digital numerals. Left: middle number can be recognized as either 6 or 8; Right: the middle row of numbers can be recognized as being aligned in either increasing or decreasing sequence. (C) After visual adaptation to normal number 6, participants recognized the test occluded character as number 8; on the other hand, after visual adaptation to normal number 8, participants recognized the test occluded character as number 6.

view more 

Credit: Junxiang Luo

Perceptual recognition of numerical characters, like Arabic numerals, is indispensable for our daily activities in the modern society. Studying the perceptual and neural mechanisms that endow us with the ability to understand those characters is an important scientific topic. In this project, researchers explored a bistable perceptual phenomenon of a specially designed character named occluded digital numeral, to get a deeper understanding of the mechanisms underpinning the perceptual recognition of numbers.
The digital numeral is a special version of number fonts designed for application in electronic products like traffic countdown timers, digital calculators, and many other LED broadcasting panels. Sometimes, an accidental partial occlusion of these characters (Fig. 1A) results in an intriguing perceptual phenomenon, that one digital numeral can be recognized as having multiple, usually two, semantic meanings (Fig. 1B).
“Because of the simple design of digital numerals, a partial occlusion causes visual ambiguities. In other words, the visual input does not contain enough information for us to make a unique semantic interpretation of the occluded numerical character,” said Dr. Junxiang Luo of the National Institute for Physiological Sciences, a researcher who led this project.


Researchers then performed a behavioral experiment on healthy human participants using visual adaptation tasks. This task paradigm examined how the exposure to specific types of visual input bias perception. Specifically, they found that after long-time presentation of normal digital numerals to the visual system, participants tended to report seeing a unique number biased away from the adaptation numeral (Fig. 1C).
“Visual adaptation selectively reduces the neural sensitivity of a certain visual processing stage to a certain category of visual stimuli, so by evaluating how much the interpretation biases away from the bistable state after adaptation to different types of visual information, we have a chance to know at which visual processing level that the perceptual recognition of numerals may happen,” said Dr. Luo.
The researchers selected a series of candidate adaptation stimuli. They demonstrated that the bistability does not depend on either low-level visual information such as simple visual features, or high-level information such as semantics, and these results suggest a contribution of midlevel visual processing that encodes complex shapes and symbolic number forms.


The perceptual phenomenon discovered in this psychophysical study has the potential to be extended to physiological studies in the future, in which researchers can compare brain activity among different perceptual interpretations of the same occluded numerals. “We hope to extend this study using some physiological methods to get a deeper understanding and more precise localization of the brain areas which generate the perceptual bistability,” said Dr. Luo.
The results are published in Journal of Vision on September 17th, 2024. Dr. Isao Yokoi and Professor Hiromasa Takemura of the National Institute for Physiological Sciences and Dr. Serge Dumoulin of the Spinoza Centre for Neuroimaging also contributed this research.

 

 


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.