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 2105.04447cs.CVcs.AI
收藏

SCTN:稀疏卷积-变换网络场景流估计

SCTN: Sparse Convolution-Transformer Network for Scene Flow Estimation

KAUST(阿卜杜拉国王科技大学)

Bing Li, Cheng Zheng, Silvio Giancola, Bernard Ghanem

摘要:

我们提出了一种新的场景流估计方法来捕获和推断点云的三维运动。估计点云的3D运动是具有挑战性的,因为点云是无序的并且其密度是显著不均匀的。这种非结构化数据在匹配点云之间的对应点时造成困难,导致不准确的流量估计。我们提出了一种新的架构称为稀疏卷积-Transformer网络(SCTN),配备了Transformer的稀疏卷积。具体而言,通过利用稀疏卷积,SCTN将不规则点云转换为局部一致的流特征,用于估计对象/局部对象部分内的连续和一致的运动。我们进一步建议显式学习点关系使用点Transformer模块,不同于现有的方法。我们表明,学习的基于关系的上下文信息是丰富的,有助于匹配对应点,有利于场景流量估计。此外,提出了一种新的损失函数,根据特征相似性自适应地激励流一致性。大量的实验表明,我们提出的方法实现了一个新的艺术状态的场景流量估计。我们的方法在FlyingThings 3D和KITTI Scene Flow上分别实现了0.038和0.037(EPE 3D)的误差,大大优于以前的方法。

英文摘要:

We propose a novel scene flow estimation approach to capture and infer 3D motions from point clouds. Estimating 3D motions for point clouds is challenging, since a point cloud is unordered and its density is significantly non-uniform. Such unstructured data poses difficulties in matching corresponding points between point clouds, leading to inaccurate flow estimation. We propose a novel architecture named Sparse Convolution-Transformer Network (SCTN) that equips the sparse convolution with the transformer. Specifically, by leveraging the sparse convolution, SCTN transfers irregular point cloud into locally consistent flow features for estimating continuous and consistent motions within an object/local object part. We further propose to explicitly learn point relations using a point transformer module, different from exiting methods. We show that the learned relation-based contextual information is rich and helpful for matching corresponding points, benefiting scene flow estimation. In addition, a novel loss function is proposed to adaptively encourage flow consistency according to feature similarity. Extensive experiments demonstrate that our proposed approach achieves a new state of the art in scene flow estimation. Our approach achieves an error of 0.038 and 0.037 (EPE3D) on FlyingThings3D and KITTI Scene Flow respectively, which significantly outperforms previous methods by large margins.

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。

↑