Abstract
Machine learning on 3D point clouds is challenging due to the absence of natural ordering of the points. PointNet is a neural network architecture capable of processing such unordered point sets directly, which has achieved promising results on classification and segmentation tasks.
We explore methods of utilizing point neighborhood features within PointNet and their impact on classification performance. We propose neural models that operate on point clouds accompanied by point features.
The results of our experiments suggest that traditional spin image representations of point neighborhoods can improve classification effectiveness of PointNet on datasets comprised of objects that are not aligned into canonical orientation. Furthermore, we introduce a feature-based alternative to spatial transformer, which is a sub-network of PointNet responsible for aligning misaligned objects into canonical orientation.
Additional experiments demonstrate that the alternative might be competitive with spatial transformer on challenging datasets.