Catching differences in visual images: performance depend on how images are presented

What is it about?

In our study, we explored human performance in detecting subtle visual differences within pairs of non-figurative Perlin-noise images, akin to a 'spot-the-difference' task. These image pairs were presented in four distinct formats: spatially separated in horizontal or vertical arrangements with simultaneous presentation, or sequentially presented with a brief delay or without any delay. In the two spatial arrangements, we observed 'change blindness'—failure to detect the target within 30 seconds—in approximately 6-7% of trials, while in the brief delay condition, this occurrence reduced to 2.4%. Notably, we achieved rapid and error-free target detection (termed 'pop-out') in the sequential format without any delay. The average detection time when the target was identified in the spatial formats was around 9 seconds, while in the brief delay condition, it improved to approximately 6 seconds. Moreover, when participants successfully detected the target, their precision in locating it was correlated in the horizontal and brief delay conditions, and superior to the vertical condition. We also observed stronger correlations between misses in the horizontal and brief delay conditions compared to misses between the vertical and horizontal, and between the vertical and brief delay conditions. Interestingly, some individuals exhibited chance-level performance when comparing images in the vertical direction, suggesting potential influences of reduced precision during saccades in the vertical versus horizontal direction. These findings may hold relevance for radiologists, as the stimuli and task bear resemblance to the challenges they face when identifying deviations in radiological images.

Featured Image

Photo by Ion Fet on Unsplash

Photo by Ion Fet on Unsplash

Why is it important?

This research is important because it delves into how humans perceive and react to subtle differences in visual information. Understanding our ability to detect changes in complex images has broader implications, including applications in fields like radiology, where identifying deviations in medical images is critical for accurate diagnoses. The study's findings shed light on the factors that influence our ability to spot these differences, such as presentation format and timing, potentially helping improve our overall visual perception and decision-making abilities in various professional and everyday contexts.