Chimera in Visual Art: Perceptual Cues, Prediction Models, and the Inference of Error

The following is an excerpt based on the presentation by Anthony Waichulis given at the 2025 IX Arts convention.

“Humans form an initial visual interpretation almost immediately upon seeing an image. The ‘gist’ of a scene can be extracted in under 100 milliseconds, with above-chance identification possible in as little as 13 milliseconds (about 1/75th of a second). By approximately 150 milliseconds after image onset, the brain has already extracted key information sufficient to differentiate basic content, such as detecting an animal or identifying a scene type. This speed may seem astonishing, but it aligns with everyday vision. Our eyes fixate on new points roughly three times per second, requiring the brain to extract meaningful information from each glimpse in under 300 milliseconds.

Even a single diagnostic object or spatial layout cue can signal a scene’s identity in as little as 50 milliseconds. During this initial feedforward sweep (a rapid, bottom-up cascade of neural activity through early brain regions), the brain detects broad patterns of shape, color, and spatial structure that suggest the kind of scene present. This enables viewers to quickly recognize whether they are viewing, for example, a beach, a cityscape, or an interior space, and to grasp the overall subject or theme of a painting almost at a glance…”

To read the full article: Chimera in Visual Art: Perceptual Cues, Prediction Models, and the Inference of Error