Triangle-Net: Towards Robustness in Point Cloud Learning
Triangle-Net: Towards Robustness in Point Cloud Learning
复制标题
DOI：
10.1109/wacv48630.2021.00087
复制
发表时间：
2020-02
期刊：
影响因子：
--
通讯作者：
Chenxi Xiao;J. Wachs
中科院分区：
文献类型：
--
作者：
Chenxi Xiao;J. Wachs
三维(3D)目标识别正在成为许多计算机视觉系统(如自动驾驶车辆、服务机器人和无人侦察机)在非结构化环境中更有效地运行所需的关键能力。这些实时系统需要有效的分类方法，这些方法对不同的采样分辨率、噪声测量和不受约束的姿势配置都是稳健的。以往的研究表明，点的稀疏性、旋转性和位置的固有方差会导致基于点云的分类技术的性能显著下降。然而，对于多因素方差和显著稀疏性，这两种方法都不够稳健。在这方面，我们提出了一种新的3D分类方法，该方法可以同时实现对旋转、位置移动、缩放的不变性，并且对点稀疏性具有很强的鲁棒性。为此，我们引入了一种新的特征，它利用点云的图形结构，可以通过我们提出的神经网络端到端地学习来获得稳健的3D对象的潜在表示。结果表明，在点稀疏的情况下，这种潜在表示可以显著提高目标分类和检索任务的性能。此外，在任意SO(3)旋转下，在仅使用16个稀疏点云的ModelNet 40分类任务中，我们的方法的性能分别比PointNet和3DmFV高35.0%和28.1%。
Three dimensional (3D) object recognition is becoming a key desired capability for many computer vision systems such as autonomous vehicles, service robots and surveillance drones to operate more effectively in unstructured environments. These real-time systems require effective classification methods that are robust to various sampling resolutions, noisy measurements, and unconstrained pose configurations. Previous research has shown that points’ sparsity, rotation and positional inherent variance can lead to a significant drop in the performance of point cloud based classification techniques. However, neither of them is sufficiently robust to multifactorial variance and significant sparsity. In this regard, we propose a novel approach for 3D classification that can simultaneously achieve invariance towards rotation, positional shift, scaling, and is robust to point sparsity. To this end, we introduce a new feature that utilizes graph structure of point clouds, which can be learned end-to-end with our proposed neural network to acquire a robust latent representation of the 3D object. We show that such latent representations can significantly improve the performance of object classification and retrieval tasks when points are sparse. Further, we show that our approach outperforms PointNet and 3DmFV by 35.0% and 28.1% respectively in ModelNet 40 classification tasks using sparse point clouds of only 16 points under arbitrary SO(3) rotation.