Abstract: To achieve high spatial resolution of X-ray imaging, the focal spot size of high-energy nondestructive testing electron accelerators has been progressively reduced from 2 to below 1 mm, and, ...
Abstract: Recent work has shown that deep vision models tend to be overly dependent on low-level or “texture” features, leading to poor generalization. Various data augmentation strategies have been ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results