A general framework for rotation invariant point cloud analysis
This work addresses rotation invariance for point cloud analysis, which is a domain-specific problem for 3D computer vision, and appears incremental as it builds on existing methods like PCA.
The authors tackled the problem of rotation sensitivity in point cloud analysis by proposing a general framework that achieves rotation invariance, showing considerable or better performance compared to state-of-the-art methods on common benchmarks.
We propose a general method for deep learning based point cloud analysis, which is invariant to rotation on the inputs. Classical methods are vulnerable to rotation, as they usually take aligned point clouds as input. Principle Component Analysis (PCA) is a practical approach to achieve rotation invariance. However, there are still some gaps between theory and practical algorithms. In this work, we present a thorough study on designing rotation invariant algorithms for point cloud analysis. We first formulate it as a permutation invariant problem, then propose a general framework which can be combined with any backbones. Our method is beneficial for further research such as 3D pre-training and multi-modal learning. Experiments show that our method has considerable or better performance compared to state-of-the-art approaches on common benchmarks. Code is available at https://github.com/luoshuqing2001/RI_framework.