Challenging the Universal Representation of Deep Models for 3D Point Cloud Registration

Learning universal representations across different applications domain is an open research problem. In fact, finding universal architecture within the same application but across different types of datasets is still unsolved problem too, especially in applications involving processing 3D point clouds. In this work we experimentally test several state-of-the-art learning-based methods for 3D point cloud registration against the proposed non-learning baseline registration method. The proposed method either outperforms or achieves comparable results w.r.t. learning based methods. In addition, we propose a dataset on which learning based methods have a hard time to generalize. Our proposed method and dataset, along with the provided experiments, can be used in further research in studying effective solutions for universal representations. Our source code is available at:

PDF Abstract


Introduced in the Paper:


Used in the Paper:

Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Point Cloud Registration ETH (trained on 3DMatch) Greedy Grid Search Recall 0.784 # 5
Point Cloud Registration FAUST-partial (60%+ overlap, Rot 0-45, Trans -50-50, trained on 3DMatch) Greedy Grid Search Recall (%) 92.81 # 1
RRE (degrees) 0.014 # 1
RTE (cm) 0.009 # 1
Point Cloud Registration KITTI (trained on 3DMatch) Greedy Grid Search Success Rate 90.27 # 3


No methods listed for this paper. Add relevant methods here