While perceptual experience appears continuous and uninterrupted, studies reveal that it is composed of periodic intervals evidenced by fluctuating responsiveness to sensory input. For instance, accuracies in simple detection tasks oscillate at 4–8 Hz (Fieblekorn, Saalmann, & Kastner, 2013), and response times in similar tasks exhibit periodic patterns at 5–15 Hz (Cha & Blake, 2024).
A compelling visual demonstration of the discrete nature of perceptual experience can be observed in the continuous Wagon Wheel Illusion (Fig. 1; VanRullen, 2006). In this demonstration, the sine wave patterns of a rotating wheel undergo periodic changes at 10 Hz, a frequency considerably slower than the frame rate of 30 Hz, which prevents temporal aliasing. Yet, we occasionally perceive the patterns on both sides rotating in the same direction. This suggests that our visual system samples sensory input at relatively low frequencies, shaping our visual processing in a rhythmic manner.
Binocular rivalry provides a valuable tool to explore this question. This phenomenon occurs when an observer is presented with very different, incompatible images to the left and right eyes (Fig. 2). In such cases, the observer perceives only one image at a time, with perception stochastically alternating between the two images.
By asking the observer to continuously report their perception during binocular rivalry, visual awareness can be probed. Interestingly, this approach reveals that transitions to the previously suppressed image occur in rhythmic cycles, with periods of frequent transitions alternating with periods of infrequent transitions (Cha & Blake, 2019). This pattern suggests that visual awareness is shaped by an underlying temporal structure.
At DCANT Lab, we explore how the brain maintains continuous and uninterrupted visual awareness despite the limitations imposed by the temporal structure of visual processing. To investigate this, we use electroencephalography (EEG) to track periodic oscillations in the brain and computational models to uncover periodicity embedded in behavior. Additionally, we develop computational models to simulate how the underlying temporal structure shapes visual awareness.
Perceptual experience often feels far richer than the brain can feasibly process at any given moment. Imagine scanning a room full of people and trying to gauge whether they are enjoying the scene. Surprisingly, this task does not feel overwhelming or impossible.
This seemingly effortless task requires extracting statistical properties from a large group of stimuli (e.g., extrating the mean emotion from a crowd of faces). Researchers suggest that we possess an ability to efficiently summarize large amounts of visual information into meaningful patterns or properties, referred to as ensemble perception (for review, see Whitney & Yamanashi Leib, 2018). For a quick demonstration, play the interactive Figure 3, wait for the two arrays of circles to flash briefly on the left and right, and silently decide which array has the larger average size. You might expect this to be a very difficult task, but you will likely find yourself agreeing with the correct response when it appears after the arrays.
At DCANT Lab, we investigate whether ensemble perception relies on shared mechanisms and how it differs from object perception (e.g., Cha, Blake, & Gauthier, 2022; Chang, Cha, & Gauthier, 2024). By studying how the brain processes large groups of stimuli, we seek to understand the mechanisms that allow visual awareness to operate seamlessly, even within the constraints of visual processing.