2022WACV论文：D2Conv3D的梳理及2D/3DCNN、可变形卷积等关键词解释

论文地址：D2Conv3D: Dynamic Dilated Convolutions for Object Segmentation in Videos

参考资料：

[1] 目标检测之Deformable Convolutional Networks(2017)：https://blog.csdn.net/P_LarT/article/details/85042264

[2] 一维卷积（1D-CNN）、二维卷积（2D-CNN）、三维卷积（3D-CNN）：https://blog.csdn.net/yizhishuixiong/article/details/106566730

[3] Schmidt C, Athar A, Mahadevan S, et al. D2Conv3D: Dynamic Dilated Convolutions for Object Segmentation in Videos[C]//Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision. 2022: 1200-1209.

[4] Mahadevan S, Athar A, Ošep A, et al. Making a case for 3d convolutions for object segmentation in videos[J]. arXiv preprint arXiv:2008.11516, 2020.

[5] Wu Y, He K. Group normalization[C]//Proceedings of the European conference on computer vision (ECCV). 2018: 3-19.