PointHop: An Explainable Machine Learning Method for Point Cloud Classification

Abstract

The PointHop method was recently proposed by Zhang et al. for 3D point cloud classification with unsupervised feature extraction. It has an extremely low training complexity while achieving stateof-the-art classification performance. In this work, we improve the PointHop method furthermore in two aspects: 1) reducing its model complexity in terms of the model parameter number and 2) ordering discriminant features automatically based on the cross-entropy criterion. The resulting method is called PointHop++. The first improvement is essential for wearable and mobile computing while the second improvement bridges statistics-based and optimization-based machine learning methodologies. With experiments conducted on the ModelNet40 benchmark dataset, we show that the PointHop++ method performs on par with deep neural network (DNN) solutions and surpasses other unsupervised feature extraction methods.

Layered Projection-Based Quality Assessment of 3D Point Clouds

“…However, this model presents relatively high computation complexity due to some operations, such as graph construction [27]. With the proposal of PC deep learning networks [34] such as PointNet [35], Liu et al [23] used a mathematical model of Artificial Neural Networks to evaluate the quality of PC. It is based on the stack of sparse convolutional layers and residuals, extracts hierarchical features, and pools them globally to obtain feature vectors.…”

PixelHop: A successive subspace learning (SSL) method for object recognition

“…2 offers a concrete example of SSL. Another design based on the SSL principle, called the PointHop method, was proposed in [14]. It is worthwhile to obtain a high-level abstraction for these methods.…”

“…This idea applies to structured data (e.g., images) as well as unstructured data (e.g., 3D point cloud sets). An SSL-based 3D point cloud classification scheme, called the PointHop method, was proposed in [14].…”

3D Point Cloud Classification for Autonomous Driving via Dense-Residual Fusion Network

“…Content may change prior to final publication. Model car pedestrians street clutter Accuracy (%) PointGCN [7] 99.00 93.00 79.50 91.00 PointNet [5] 98.00 85.00 45.00 76.00 PointHop [9] 99 PointNet++ [6], PointHop [9], DGCNN [8], and Chiang et al [13]. The model by Chiang et al [13] projects the point clouds into BA images, while those of DGCNN and PointGCN convert the point clouds into graph signals.…”

“…PointHop [9] adopts k-nearest neighbors to group points in the point cloud. The points in the same group are divided into eight octants around the group center.…”