equivariant专题

ReDet:A Rotation-equivariant Detector for Aerial Object Detection

ReDet:A Rotation-equivariant Detector for Aerial Object Detection 文章目录 ReDet:A Rotation-equivariant Detector for Aerial Object Detection参考IntroductionRelated WorksPreliminariesRotation-equivaria

Equivariant GNN Networks(等变图神经网络)

3D CNN的缺陷:  旋转不等变、很难扩展到大的蛋白质口袋 The model was constructed based on the 3D CNN model which is not equivariant to rotation and hard to scale to large protein pockets. Vector Feature-based Equivari

[论文理解]旋转等变向量场网络Rotation equivariant vector field networks

1. 几个问题 做了什么 提出了旋转等变向量场网络Rotation Equivariant Vector Field Networks (RotEqNet),是一种对旋转等变性(equivariance)、旋转不变性(invariance)和旋转共变性(covariance)进行了编码的CNN网络。 论文中专注于解决“旋转”这一操作,因此下文所述的对象都与旋转有关。 实现原理 每个卷