SPE-Net: Boosting Point Cloud Analysis via Rotation Robustness Enhancement

TL;DR

SPE-Net is a novel deep learning architecture for 3D point cloud analysis that enhances rotation robustness by encoding rotation information through an attention-based mechanism, improving performance on rotated and unrotated data.

Contribution

The paper introduces SPE-Net, a new architecture with Selective Position Encoding that effectively leverages rotation information to improve robustness in point cloud analysis.

Findings

SPE-Net outperforms state-of-the-art methods on multiple benchmarks.

Encoding local rotation information benefits test accuracy even without global rotation.

SPE-Net demonstrates strong robustness to rotated point cloud data.

Abstract

In this paper, we propose a novel deep architecture tailored for 3D point cloud applications, named as SPE-Net. The embedded ``Selective Position Encoding (SPE)'' procedure relies on an attention mechanism that can effectively attend to the underlying rotation condition of the input. Such encoded rotation condition then determines which part of the network parameters to be focused on, and is shown to efficiently help reduce the degree of freedom of the optimization during training. This mechanism henceforth can better leverage the rotation augmentations through reduced training difficulties, making SPE-Net robust against rotated data both during training and testing. The new findings in our paper also urge us to rethink the relationship between the extracted rotation information and the actual test accuracy. Intriguingly, we reveal evidences that by locally encoding the rotation information through SPE-Net, the rotation-invariant features are still of critical importance in benefiting the test samples without any actual global rotation. We empirically demonstrate the merits of the SPE-Net and the associated hypothesis on four benchmarks, showing evident improvements on both rotated and unrotated test data over SOTA methods. Source code is available at https://github.com/ZhaofanQiu/SPE-Net.