Evaluating Synthetic Images as Effective Substitutes for Experimental Data in Surface Roughness Classification
Authors:
Binwei Chen,
Huachao Leng,
Chi Yeung Mang,
Tsz Wai Cheung,
Yanhua Chen,
Wai Keung Anthony Loh,
Chi Ho Wong,
Chak Yin Tang
Abstract:
Hard coatings play a critical role in industry, with ceramic materials offering outstanding hardness and thermal stability for applications that demand superior mechanical performance. However, deploying artificial intelligence (AI) for surface roughness classification is often constrained by the need for large labeled datasets and costly high-resolution imaging equipment. In this study, we explor…
▽ More
Hard coatings play a critical role in industry, with ceramic materials offering outstanding hardness and thermal stability for applications that demand superior mechanical performance. However, deploying artificial intelligence (AI) for surface roughness classification is often constrained by the need for large labeled datasets and costly high-resolution imaging equipment. In this study, we explore the use of synthetic images, generated with Stable Diffusion XL, as an efficient alternative or supplement to experimentally acquired data for classifying ceramic surface roughness. We show that augmenting authentic datasets with generative images yields test accuracies comparable to those obtained using exclusively experimental images, demonstrating that synthetic images effectively reproduce the structural features necessary for classification. We further assess method robustness by systematically varying key training hyperparameters (epoch count, batch size, and learning rate), and identify configurations that preserve performance while reducing data requirements. Our results indicate that generative AI can substantially improve data efficiency and reliability in materials-image classification workflows, offering a practical route to lower experimental cost, accelerate model development, and expand AI applicability in materials engineering.
△ Less
Submitted 26 March, 2026;
originally announced March 2026.
Exotic Pressure-Driven Band Gap Widening in Carbon Chain-Filled KFI Zeolite and Its Pathway to High-Pressure Semiconducting Electronics and High-Temperature Superconductivity
Authors:
C. T. Wat,
K. C. Lam,
W. Y. Chan,
C. P. Chau,
S. P. Ng,
W. K. Loh,
L. Y. F. Lam,
X. Hu,
C. H. Wong
Abstract:
Semiconducting devices face persistent challenges in operating at high pressure, as the band theory predicts that materials transition to a more metallic state under compression. However, our findings with carbon chains in KFI substrates reveal a conditional deviation from this norm. We not only witness the transition from polyyne (semiconductor) to cumulene (metal) at medium pressure, but we also…
▽ More
Semiconducting devices face persistent challenges in operating at high pressure, as the band theory predicts that materials transition to a more metallic state under compression. However, our findings with carbon chains in KFI substrates reveal a conditional deviation from this norm. We not only witness the transition from polyyne (semiconductor) to cumulene (metal) at medium pressure, but we also observe an unexpected re-entrance of the polyyne at high pressures, where the band gap in the polyyne increases with pressure. In addition, the synthesis of long cumulene chains has posed a longstanding challenge in the quest for high-temperature organic superconductivity. We have identified critical conditions for synthesizing extended cumulene chains within zeolite frameworks, highlighting the interplay between unconventional charge density waves and significant torsions. The KFI zeolite facilitates the formation of carbon chains exceeding 5,000 atoms, in stark contrast to around 100 other zeolites that are limited to ~10 atoms. The cumulene@KFI system demonstrates a superconducting transition temperature reaching ~62 K, surpassing the highest reported values for bulk iron-based superconductors. This interplay between carbon structures and superconductivity not only advances our understanding of charge density waves but also heralds a new era in the study of novel applications
△ Less
Submitted 6 March, 2026;
originally announced March 2026.