DeepHomographyEstimation
这是基于深度学习神经网络预测单应性矩阵,并用于后续拼接和其他视觉任务的论文汇总。其中涉及的论文有如下几篇。所有论文的数据集整理在dataset文件夹下。网络的结构、损失函数的设计以及论文各自的方法将总结在doc文件夹下,点击以下的超链接直接跳转。后续会持续更新总结相关论文的思路方法。
基于监督学习的单应估计网络
- Deep image homography estimation(2016)
- A view-free image stitching network based on global homography(2020)
- Real-Time Image Stitching with Convolutional Neural Networks(2019)
- Learning Edge-Preserved Image Stitching from Large-Baseline Deep Homography(2020)
基于非监督学习的单应估计网络
- Unsupervised Deep Homography: A Fast and Robust Homography Estimation Model(2018)
- Content-Aware Unsupervised Deep Homography Estimation(2020)
- Depth-Aware Multi-Grid Deep Homography Estimation with Contextual Correlation(2021)
- DeepMeshFlow: Content Adaptive Mesh Deformation for Robust Image Registration
数据集汇总
- Synthetic Dataset(COCO合成数据集)
- Content-Aware-DeepH-Data(视频数据集)
- UDIS-D(视差宽基线图像)
!- 文章中公式不渲染的部分建议下载chrome插件MathJax Plugin for Github