
With the proliferation of text-to-image generative AI, understanding the fidelity of their output is critical. While these models can generate visually stunning images, their interpretation of nuanced, subjective concepts like color names remains largely unquantified. This paper introduces a systematic framework to evaluate how accurately leading generative AI models (including Flux, Ideogram, Kandinsky, Gemini and Stable Diffusion) understand and reproduce colors from textual prompts. We prompted these models with both one-word (e.g., ”blue”) and two-word (e.g., ”sky blue”) color names to generate uniform color fields. The resulting images were analyzed by converting them to the perceptually uniform CIE Lab color space. An adaptive k-means clustering algorithm was employed to extract the dominant color, mitigating issues of non-uniformity in the generated images. By calculating the perceptual color difference using CIEDE2000 (ΔE00) and the chromatic distance (Δab) between the AI-generated colors and standardized ground-truth values, we provide a quantitative benchmark of each model’s color accuracy. Our findings reveal that while all models broadly understand the mapping between color names and hue, significant performance variations exist among models, with systematic differences in lightness and chroma reproduction. Per-model analysis reveals a clear hierarchy in chromatic fidelity: Gemini and Flux demonstrate the strongest anchoring, while Kandinsky exhibits striking hue-dependent anisotropy and Stable Diffusion shows the broadest isotropic dispersion. Per-color analysis identifies systematic undersaturation of short-wavelength and high-chroma colors (blue, indigo, magenta) across all models, while warm colors (red, orange, yellow) are generally better grounded. We highlight that results vary significantly across random seeds for the same prompt and model, and that lexical specificity generally—but not universally—improves chromatic grounding. This work provides a robust methodology for auditing and improving color fidelity in future generative models.

Metamer mismatch bodies (MMBs) quantify the extent of metamer mismatching for a given color stimulus with a change in color mechanism (i.e. change in illuminant and/or observer). Prior work has shown that the MMB boundary can be efficiently approximated by spherical sampling of unit directions in the 6D joint color-mechanism space, and for each sampled direction, maximizing the boundary point subject to the metameric cross-section constraints. Many sampled directions map to the same boundary vertex, so the number of recovered vertices is typically far smaller than the number of sampled directions. This produces a plausible approximation, but the resulting boundary vertices, expressed in sensor-response spaces (for example XYZ, LMS, or RGB), are often distributed in a highly non-uniform manner. Increasing the number of sampled directions increases the number of recovered vertices but does not improve boundary uniformity. We explored a simple post-processing workflow that builds a larger candidate pool of vertices and then selects a fixed-size subset using a spacing-driven sampling algorithm, improving vertex uniformity as measured by a nearest-neighbor metric. This approach substantially improves vertex uniformity in sensor space, but it can discard boundary-defining extreme vertices, potentially altering hull volume and other distinguishing boundary features. We therefore argue that any practical workflow for improving MMB vertex uniformity should include an explicit mechanism for retaining boundary-critical extremes prior to applying spacing-driven selection.

Teaching color science to Electrical Engineering and Computer Science (EECS) students is critical to preparing them for advanced topics such as graphics, visualization, imaging, Augmented/Virtual Reality. Color historically receive little attention in EECS curriculum; students find it difficult to grasp basic concepts. This is because today's pedagogical approaches are non-intuitive and lack rigor for teaching color science. We develop a set of interactive tutorials that teach color science to EECS students. Each tutorial is backed up by a mathematically rigorous narrative, but is presented in a form that invites students to participate in developing each concept on their own through visualization tools. This paper describes the tutorial series we developed and discusses the design decisions we made.

There are not many international courses that teach color science applied to non-technical fields. Here, we want to present the experience from a master course organized by Gruppo del Colore – Associazione Italiana Colore and Politecnico of Milano: the Master in Color Design & Technology. It has the aim of training students in the use of color mainly for design purposes, and color science and colorimetry have a big role in the program. The Master is organized in three main phases: Fundamentals, Project Works and Internship. The first part is a series of lectures and frontal lessons that gives to the students the theorical and technical bases to be used in project works practical applications of and internship. In fact, in the second and third parts students are asked to design and develop different applicative design projects. After the first part of Fundamentals training, students will be able to manage disciplines such as colorimetry, visual perception, physics, chemistry, optics and psychology, to build up their future professional career in different fields.