Abstract: Small sample image classification is an important issue in the field of computer vision, and it has a wide application prospect in resource-limited edge devices. Therefore, this paper ...

TOKYO—Imagine a sheet made of microscopic glass fibers, woven by a former silk maker and thinner than a human hair. A shortage of this material—essential in artificial-intelligence chips—is looming ...

Abstract: We propose a low-shot image classification method called Limo, which can train an accurate image classification model under conditions of acute data scarcity. Limo uniquely assembles ...