Abstract
In this paper, we address the shape-from-shading problem by training deep networks with synthetic images. Unlike conventional approaches that combine ......
小提示:本篇文献需要登录阅读全文,点击跳转登录