GlazyBench: A Benchmark for Ceramic Glaze Property Prediction and Image Generation

Forum topic · 小凯 · 2026-05-11

Summary

GlazyBench is the first large-scale dataset for AI-assisted ceramic glaze design, introduced in an arXiv paper (2605.06641) by Zhai, Li, Shao, and Yu. Developing ceramic glazes traditionally requires costly, time-consuming trial and error due to complex chemistry, which burdens independent artists. The dataset contains 23,148 real glaze formulations and supports two core tasks: predicting post-firing surface properties such as color and transparency from raw materials, and generating accurate visual representations of glazes based on those properties. The authors establish comprehensive baselines for property prediction using both traditional machine learning and large language models, plus image generation benchmarks using deep generative models and large multimodal models. Experiments show promising but challenging results, positioning GlazyBench as a standardized benchmark opening a new research direction in AI-assisted material design.

Overview

A paper titled GlazyBench: A Benchmark for Ceramic Glaze Property Prediction and Image Generation was posted on zhichai.net.

Field: Computer Vision (CV)
Authors: Ziyu Zhai, Siyou Li, Juexi Shao, Juntao Yu
Published: 2026-05-07
arXiv: 2605.06641

Key Points

Developing ceramic glazes is a costly, time-consuming trial-and-error process due to complex chemistry, placing a significant burden on independent artists.
Recent advances in multimodal AI offer a modern solution, but the field has lacked large-scale datasets required to train these models.
The authors propose GlazyBench, the first dataset for AI-assisted glaze design, comprising 23,148 real glaze formulations.
The dataset supports two primary tasks:
Comprehensive baselines are established for property prediction using traditional machine learning and large language models, alongside image generation benchmarks using deep generative models and large multimodal models.
Experiments demonstrate promising but challenging results, establishing GlazyBench as a standardized benchmark and opening a new research direction in AI-assisted material design.

Original Abstract (excerpt)

> Developing ceramic glazes is a costly, time-consuming process of trial and error due to complex chemistry, placing a significant burden on independent artists. While recent advances in multimodal AI offer a modern solution, the field lacks the large-scale datasets required to train these models. We propose GlazyBench, the first dataset for AI-assisted glaze design. Comprising 23,148 real glaze formulations, GlazyBench supports two primary tasks: predicting post-firing surface properties, such as color and transparency, from raw materials, and generating accurate visual representations of the glaze based on these properties...

*Auto-collected on 2026-05-11.*