In short: color and space aren’t processed in isolation. They’re intertwined in large-scale, region-specific maps that generalize across humans. This opens questions about adaptation of color processing possibly to optimize object- and scene perception in natural vision.

Why does this matter? It suggests functional or evolutionary pressures have organized color coding in a spatially structured, conserved way—hinting at deep links between how we perceive where things are and what color they are.

Those region-specific biases were shared across brains. Meaning the way your V1 vs. V4 links space and color looks remarkably like mine, despite individual variability.

This worked across multiple visual areas (V1–V3, hV4, LO1-2). Importantly, each area showed its own idiosyncratic color biases across retinotopic space. So different regions map color onto space in their own unique—but consistent—ways.

Then we tested whether we could predict what color someone was seeing based only on brain activity patterns from other people’s brains. A classifier trained on others’ color responses could decode colors in a new subject.

We used fMRI and a clever trick: align brains based only on responses to achromatic spatial patterns used for retinotopic mapping across individuals, without using any color information.

The questions we asked: does the same color trigger comparable neural activity across different people? And do brain regions encode colors in distinct, systematic ways? We set out to test this directly.

The human brain’s color responses are similar across individuals, forming large-scale, region-specific maps.