Extreme Rotation Estimation using Dense Correlation Volumes
Extreme Rotation Estimation using Dense Correlation Volumes
复制标题
DOI：
10.1109/cvpr46437.2021.01433
复制
发表时间：
2021-04
期刊：
影响因子：
--
通讯作者：
Ruojin Cai;Bharath Hariharan;Noah Snavely;Hadar Averbuch-Elor
中科院分区：
文献类型：
--
作者：
Ruojin Cai;Bharath Hariharan;Noah Snavely;Hadar Averbuch-Elor
我们提出了一种技术，用于估计的RGB图像对在极端的设置，其中的图像有很少或没有重叠的相对3D旋转。我们观察到，即使图像不重叠，也可能有丰富的隐藏线索，它们的几何关系，如光源方向，消失点，和对称性存在于场景中。我们提出了一个网络设计，可以通过比较两个输入图像之间的所有点对来自动学习这种隐式提示。因此，我们的方法构建密集的特征相关体积，并处理这些来预测相对的3D旋转。我们的预测是在旋转的细粒度离散化上形成的，绕过了与回归3D旋转相关的困难。我们在各种极端RGB图像对上展示了我们的方法，包括在不同光照条件和地理位置下捕获的室内和室外图像。我们的评估表明，我们的模型可以成功地估计非重叠图像之间的相对旋转，而不会影响重叠图像对的性能。
We present a technique for estimating the relative 3D rotation of an RGB image pair in an extreme setting, where the images have little or no overlap. We observe that, even when images do not overlap, there may be rich hidden cues as to their geometric relationship, such as light source directions, vanishing points, and symmetries present in the scene. We propose a network design that can automatically learn such implicit cues by comparing all pairs of points between the two input images. Our method therefore constructs dense feature correlation volumes and processes these to predict relative 3D rotations. Our predictions are formed over a fine-grained discretization of rotations, bypassing difficulties associated with regressing 3D rotations. We demonstrate our approach on a large variety of extreme RGB image pairs, including indoor and outdoor images captured under different lighting conditions and geographic locations. Our evaluation shows that our model can successfully estimate relative rotations among non-overlapping images without compromising performance over overlapping image pairs.1