论文标题
End2END多视图功能与可区分姿势优化匹配
End2End Multi-View Feature Matching with Differentiable Pose Optimization
论文作者
论文摘要
错误的特征匹配对随后的相机姿势估计有严重影响，并且通常需要采取额外的时间，例如RANSAC，例如兰萨克（Ransac）来进行异常拒绝。我们的方法通过共同解决特征匹配和姿势优化来应对这一挑战。为此，我们提出了一个图形注意网络，以预测图像对应关系以及置信度。在可区分的姿势估计中，结果匹配是加权约束。训练功能与姿势优化的梯度匹配自然地学会了与Superglue在扫描板上的距离和姿势估算的下降异常值，并提高了图像对的姿势估计。同时，它将姿势估计时间缩短了50％以上，而不必要的抢劫弹药。此外，我们通过跨多个框架跨越图形来一次预测匹配项，从多个视图中整合了信息。与Superglue相比，多视图匹配与端到端培训相结合，将Matterport3d的姿势估计指标提高了18.5％。
Erroneous feature matches have severe impact on subsequent camera pose estimation and often require additional, time-costly measures, like RANSAC, for outlier rejection. Our method tackles this challenge by addressing feature matching and pose optimization jointly. To this end, we propose a graph attention network to predict image correspondences along with confidence weights. The resulting matches serve as weighted constraints in a differentiable pose estimation. Training feature matching with gradients from pose optimization naturally learns to down-weight outliers and boosts pose estimation on image pairs compared to SuperGlue by 6.7% on ScanNet. At the same time, it reduces the pose estimation time by over 50% and renders RANSAC iterations unnecessary. Moreover, we integrate information from multiple views by spanning the graph across multiple frames to predict the matches all at once. Multi-view matching combined with end-to-end training improves the pose estimation metrics on Matterport3D by 18.5% compared to SuperGlue.
