PointHop++: A Lightweight Learning Model on Point Sets for 3D Classification
This work addresses lightweight 3D classification for wearable and mobile computing, but it is incremental as it builds directly on the existing PointHop method.
The authors tackled the problem of 3D point cloud classification by improving the PointHop method to reduce model complexity and automatically order discriminant features, resulting in PointHop++ which performs on par with deep neural networks and surpasses other unsupervised methods on the ModelNet40 dataset.
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 state-of-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.