The Temporal Dynamics of Ensemble Perception

The striking disparity between the subjective richness of experience and the considerable limitations of perceptual processing has emerged as an essential, enduring question in both vision science and philosophy of mind. A potential solution to this issue is ensemble perception: the ability for the visual system to compute the statistical summaries of object groups, effectively compressing an otherwise overwhelming amount of information. Previous work has supported that ensemble statistics can be perceived quickly and accurately for a wide range of object features. This has motivated models of ensemble perception as an early process in vision, providing an initial sense of the visual world and helping guide limited resources to areas of interest. These models offer compelling evidence that ensemble perception imbues a richness both to our initial views of a complex scenes, as well as to the ongoing representation of information outside the focus of attention. However, the precise temporal dynamics of ensemble perception remains a critical area of study, to both confirm and expand upon these models. My dissertation approaches this task with two primary aims: testing the initial timing of ensemble perception in vision and investigating how it operates dynamically across time. I begin by introducing and reviewing the current understanding of ensemble perception, outlining how my research is situated within, and can help to advance contemporary theories of perception. In Chapter Two, I present results from my EEG study demonstrating that ensemble statistics can be processed more rapidly than individual object details. Chapter Three discusses a series of behavioral experiments supporting that ensemble perception operates as an iteratively updating process, cleaning and revising statistical representations over time. In Chapter Four, I use fMRI data to demonstrate that judging the ensemble properties of groups evokes clearly distinguishable activity in areas such as V2 and V3. These results support and expand upon evidence of ensemble perception as an early visual process by further associating it with activity in specifically feature responsive areas of visual cortex. Finally, in Chapter Five, I explore how these results together contribute to building robust models of ensemble perception as an early and continuously updating process, and thus advance broader theories of ensemble perception in vision. Together, this work significantly advances our understanding of the temporal dynamics of ensemble perception, laying important groundwork for further investigation into the complexities of information processing in the visual system