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

TL;DR

GlazyBench is a large-scale dataset for AI-assisted ceramic glaze design, enabling property prediction and image generation to streamline the complex and costly process of developing new glazes.

Contribution

It introduces the first comprehensive dataset for ceramic glaze property prediction and image generation, supporting multiple AI tasks and establishing baseline benchmarks.

Findings

Promising results in property prediction using traditional ML and LLMs.

Effective image generation benchmarks with deep generative and multimodal models.

Demonstrates the potential and challenges of AI in ceramic glaze design.

Abstract

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. We establish comprehensive baselines for property prediction using traditional machine learning and large language models, alongside image generation benchmarks using deep generative and large multimodal models. Our experiments demonstrate promising yet challenging results. GlazyBench pioneers a new research direction in AI-assisted material design, providing a standardized benchmark for systematic evaluation.