Visual Noise from Natural Scene Statistics Reveals Human Scene Category Representations

Our visual perceptions are guided both by the bottom-­‐up information entering our eyes as well as our top-­‐down expectations of what we will see. Although bottom-­‐up visual processing has been extensively studied, comparatively little is known about top-­‐down signals. Here, we describe REVEAL (Representations Envisioned Via Evolutionary ALgorithm), a method for visualizing an observer's internal representation of a complex, real-­‐world scene, allowing us to visualize top-­‐down visual information. REVEAL rests on two innovations for solving this high-­‐ dimensional problem: visual noise that samples from natural image statistics, and a computer algorithm that collaborates with human observers to efficiently obtain a solution. In this work, we visualize observers' internal representations of a visual scene category (street) using an experiment in which the observer views visual noise and collaborates the algorithm to recreate his internal representation. As no scene information was presented, observers had to use their internal knowledge of the target, matching it with the visual features in the noise. We demonstrate that observers can use this method to re-­‐create a specific photograph, as well as to visualize purely mental images. Critically, we show that the visualized mental images can be used to predict rapid scene detection performance, as each observer had faster and more accurate responses in detecting real-­‐world images that were similar to his template. These results show that it is possible to visualize previously unobservable mental representations of real world stimuli. More broadly, REVEAL provides a general method for objectively examining the content of subjective mental experiences.