Deep Models with Fusion Strategies for MVP Point Cloud Registration
Lifa Zhu, Changwei Lin, Dongrui Liu, Xin Li, Francisco Gómez-Fernández
2021-10-18Point Cloud Registration
Abstract
The main goal of point cloud registration in Multi-View Partial (MVP) Challenge 2021 is to estimate a rigid transformation to align a point cloud pair. The pairs in this competition have the characteristics of low overlap, non-uniform density, unrestricted rotations and ambiguity, which pose a huge challenge to the registration task. In this report, we introduce our solution to the registration task, which fuses two deep learning models: ROPNet and PREDATOR, with customized ensemble strategies. Finally, we achieved the second place in the registration track with 2.96546, 0.02632 and 0.07808 under the the metrics of Rot\_Error, Trans\_Error and MSE, respectively.
Related Papers
A Multi-Level Similarity Approach for Single-View Object Grasping: Matching, Planning, and Fine-Tuning2025-07-16Simultaneous Localization and Mapping Using Active mmWave Sensing in 5G NR2025-07-07CA-I2P: Channel-Adaptive Registration Network with Global Optimal Selection2025-06-26Correspondence-Free Multiview Point Cloud Registration via Depth-Guided Joint Optimisation2025-06-18MT-PCR: A Hybrid Mamba-Transformer with Spatial Serialization for Hierarchical Point Cloud Registration2025-06-16Robust Filtering -- Novel Statistical Learning and Inference Algorithms with Applications2025-06-13Rectified Point Flow: Generic Point Cloud Pose Estimation2025-06-05Cross-modal feature fusion for robust point cloud registration with ambiguous geometry2025-05-19