arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2104.02538cs.CV
收藏

基于图神经网络和相对位姿监控的视觉摄像机重定位

Visual Camera Re-Localization Using Graph Neural Networks and Relative Pose Supervision

ETH Zurich(苏黎世联邦理工学院);Niantic;University College London(伦敦大学学院)

Mehmet Ozgur Turkoglu, Eric Brachmann, Konrad Schindler, Gabriel Brostow, Aron Monszpart

摘要:

视觉重新定位意味着使用单个图像作为输入来估计相机相对于预先记录的环境的位置和方向。得分最高的方法是“基于结构的”,并且需要查询相机的内在函数作为模型的输入,并进行仔细的几何优化。当缺少内在函数时,方法通过做出各种其他假设来竞争准确性。这会产生相当好的本地化分数,但模型在某些方面是“狭窄的”,例如,需要昂贵的测试时计算、深度传感器或多个查询帧。相比之下,我们提出的方法几乎没有特殊的假设,并且在训练和测试中相当轻量级。我们的姿势回归网络只从训练场景的相对姿势中学习。为了进行推理,它构建了一个将查询图像连接到训练对象的图,并使用了一个图神经网络(GNN),其中节点上有图像表示,边缘上有图像对表示。通过在它们之间有效地传递消息,两种表示类型都被细化以产生一致的相机姿态估计。我们验证了我们的方法在标准室内(7场景)和室外(剑桥地标)相机重新定位基准的有效性。我们的相对姿势回归方法与绝对姿势回归网络的准确性相匹配,同时保留了相对姿势模型的测试时间速度和推广到非训练场景的能力。

英文摘要:

Visual re-localization means using a single image as input to estimate the camera's location and orientation relative to a pre-recorded environment. The highest-scoring methods are "structure based," and need the query camera's intrinsics as an input to the model, with careful geometric optimization. When intrinsics are absent, methods vie for accuracy by making various other assumptions. This yields fairly good localization scores, but the models are "narrow" in some way, eg., requiring costly test-time computations, or depth sensors, or multiple query frames. In contrast, our proposed method makes few special assumptions, and is fairly lightweight in training and testing. Our pose regression network learns from only relative poses of training scenes. For inference, it builds a graph connecting the query image to training counterparts and uses a graph neural network (GNN) with image representations on nodes and image-pair representations on edges. By efficiently passing messages between them, both representation types are refined to produce a consistent camera pose estimate. We validate the effectiveness of our approach on both standard indoor (7-Scenes) and outdoor (Cambridge Landmarks) camera re-localization benchmarks. Our relative pose regression method matches the accuracy of absolute pose regression networks, while retaining the relative-pose models' test-time speed and ability to generalize to non-training scenes.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑