Real3D-Aug: Point Cloud Augmentation by Placing Real Objects with Occlusion Handling for 3D Detection and Segmentation
This work addresses the annotation cost issue for 3D detection and segmentation tasks, offering an incremental improvement over existing methods.
The paper tackles the problem of expensive annotation for 3D lidar point cloud data by proposing a data augmentation method that reuses annotated real objects with occlusion handling, resulting in significant performance gains such as a 6.65% average precision improvement for the 'Hard' pedestrian class in KITTI object detection.
Object detection and semantic segmentation with the 3D lidar point cloud data require expensive annotation. We propose a data augmentation method that takes advantage of already annotated data multiple times. We propose an augmentation framework that reuses real data, automatically finds suitable placements in the scene to be augmented, and handles occlusions explicitly. Due to the usage of the real data, the scan points of newly inserted objects in augmentation sustain the physical characteristics of the lidar, such as intensity and raydrop. The pipeline proves competitive in training top-performing models for 3D object detection and semantic segmentation. The new augmentation provides a significant performance gain in rare and essential classes, notably 6.65% average precision gain for "Hard" pedestrian class in KITTI object detection or 2.14 mean IoU gain in the SemanticKITTI segmentation challenge over the state of the art.