Helmlab : A Data-Driven Analytical Color Space for UI Design Systems
Görkem Yıldız · 2026
Read Helmlab : A Data-Driven Analytical Color Space for UI Design Systems by Görkem Yıldız (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-23010.txt.
Description
Gorkem Yildiz presents Helmlab, a family of two purpose-built color spaces for UI design systems sharing an eleven-stage analytical structure. MetricSpace, a 72-parameter space, targets color-difference prediction, while GenSpace, a 44-parameter space, targets gradient and palette generation. The forward transform maps CIE XYZ to a perceptual Lab form through learned matrices, per-channel power compression, Fourier hue correction, and Helmholtz-Kohlrausch lightness adjustment.
On COMBVD, about thirty-eight hundred color pairs, MetricSpace scores about twenty-two and a half on STRESS, a twenty-three percent cut from CIEDE2000 near twenty-nine. It also leads on a screen-condition set, about twenty-three versus about sixty-three, though it trails CIEDE2000 on a small 3D-printed set, a regression the author owns. GenSpace wins 65 of 90 gradient and palette metrics against OKLab, and the transform inverts with round-trip errors below one part in ten trillion.
Catalog record
| CREATOR | Görkem Yıldız |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 28 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-23010