UOD: Unseen Object Detection in 3D Point Cloud
This addresses a critical challenge in autonomous driving for detecting unknown objects in the wild, though it appears incremental as it enhances existing methods rather than introducing a new paradigm.
The paper tackles the problem of detecting unseen 3D objects in point clouds, crucial for autonomous driving, by proposing methods like anomaly sample augmentation and learning universal objectness, resulting in large performance gains across benchmarks such as KITTI Misc and synthetic OOD datasets.
Existing 3D object detectors encounter extreme challenges in localizing unseen 3D objects and recognizing them as unseen, which is a crucial technology in autonomous driving in the wild. To address these challenges, we propose practical methods to enhance the performance of 3D detection and Out-Of-Distribution (OOD) classification for unseen objects. The proposed methods include anomaly sample augmentation, learning of universal objectness, learning of detecting unseen objects, and learning of distinguishing unseen objects. To demonstrate the effectiveness of our approach, we propose the KITTI Misc benchmark and two additional synthetic OOD benchmarks: the Nuscenes OOD benchmark and the SUN-RGBD OOD benchmark. The proposed methods consistently enhance performance by a large margin across all existing methods, giving insight for future work on unseen 3D object detection in the wild.