Multi-Scale Neighborhood Occupancy Masked Autoencoder for Self-Supervised Learning in LiDAR Point Clouds
This work addresses a critical bottleneck for self-supervised learning in 3D point cloud perception for autonomous vehicles, offering a flexible method that improves performance on key tasks.
The paper tackled the challenge of applying masked autoencoders to LiDAR point clouds in automated driving, where large empty areas cause information leakage and high computational costs, by proposing a neighborhood occupancy MAE (NOMAE) that reconstructs occupancy only near non-masked voxels at multiple scales, achieving new state-of-the-art results on nuScenes and Waymo Open datasets for semantic segmentation and 3D object detection.
Masked autoencoders (MAE) have shown tremendous potential for self-supervised learning (SSL) in vision and beyond. However, point clouds from LiDARs used in automated driving are particularly challenging for MAEs since large areas of the 3D volume are empty. Consequently, existing work suffers from leaking occupancy information into the decoder and has significant computational complexity, thereby limiting the SSL pre-training to only 2D bird's eye view encoders in practice. In this work, we propose the novel neighborhood occupancy MAE (NOMAE) that overcomes the aforementioned challenges by employing masked occupancy reconstruction only in the neighborhood of non-masked voxels. We incorporate voxel masking and occupancy reconstruction at multiple scales with our proposed hierarchical mask generation technique to capture features of objects of different sizes in the point cloud. NOMAEs are extremely flexible and can be directly employed for SSL in existing 3D architectures. We perform extensive evaluations on the nuScenes and Waymo Open datasets for the downstream perception tasks of semantic segmentation and 3D object detection, comparing with both discriminative and generative SSL methods. The results demonstrate that NOMAE sets the new state-of-the-art on multiple benchmarks for multiple point cloud perception tasks.