Robust Point Cloud Processing
through Positional Embedding
1 The University of Adelaide
2 Amazon
Illustration of different point cloud processing architectures.
SG: sampling and grouping.
PE: positional encoding.
In per-point embedding (PPE), light blue PPEs are trained end-to-end while purple ones are randomly initialized.
Both PointNet-based and PCT-based methods process points with a per-point encoder first and then use a pooling (usually max-pool) to get a feature vector for the point cloud.
We found that all the pre-processing stages in PCT could be replaced by a simple PE (our PE-AT model) while maintaining good robustness to OOD noise.
Furthermore, we can use PE as a non-learned PPE (our model) and achieves comparable performance.
A pooling operation is used to get the global feature vector.
And all these models can be used in various downstream tasks, such as classification.
Abstract
End-to-end trained per-point embeddings are an essential ingredient of any state-of-the-art 3D point cloud processing such as detection or alignment.
Methods like PointNet, or the more recent point cloud transformer---and its variants---all employ learned per-point embeddings.
Despite impressive performance, such approaches are sensitive to out-of-distribution (OOD) noise and outliers.
In this paper, we explore the role of an analytical per-point embedding based on the criterion of bandwidth.
The concept of bandwidth enables us to draw connections with an alternate per-point embedding---positional embedding, particularly random Fourier features.
We present compelling robust results across downstream tasks such as point cloud classification and registration with several categories of OOD noise.
Citation
@misc{zheng2023robust,
title={Robust Point Cloud Processing through Positional Embedding},
author={Jianqiao Zheng and Xueqian Li and Sameera Ramasinghe and Simon Lucey},
year={2023},
eprint={2309.00339},
archivePrefix={arXiv},
primaryClass={cs.CV}
}