breeze_blows的博客

论文链接：https://arxiv.org/pdf/1809.02983.pdfCVPR2019的一篇文章 代码链接：https://github.com/junfu1115/DANet/ 主要思想：按照文中的说法就是we append two types of attention modules on top of dilated FCN, which model the semantic interdependencies in spatial and channel dimensions re.

论文链接：https://arxiv.org/pdf/1909.13226.pdfCVPR2020oral 代码链接：https://github.com/xieenze/PolarMask

we propose a simple, quantization-free layer, called RoIAlign, that faithfully preserves exact spatial locations. Despite being a seemingly minor changeSecond, we found it essential to decouple mask ...