BEVDepth: Acquisition of Reliable Depth for Multi-view 3D Object Detection
It improves multi-view 3D object detection for autonomous driving by addressing a key bottleneck in depth estimation, though it is incremental in method.
The paper tackles the problem of unreliable depth estimation in camera-based 3D object detection by proposing BEVDepth, which uses explicit depth supervision and a depth refinement module, achieving a state-of-the-art 60.9% NDS on the nuScenes dataset.
In this research, we propose a new 3D object detector with a trustworthy depth estimation, dubbed BEVDepth, for camera-based Bird's-Eye-View (BEV) 3D object detection. Our work is based on a key observation -- depth estimation in recent approaches is surprisingly inadequate given the fact that depth is essential to camera 3D detection. Our BEVDepth resolves this by leveraging explicit depth supervision. A camera-awareness depth estimation module is also introduced to facilitate the depth predicting capability. Besides, we design a novel Depth Refinement Module to counter the side effects carried by imprecise feature unprojection. Aided by customized Efficient Voxel Pooling and multi-frame mechanism, BEVDepth achieves the new state-of-the-art 60.9% NDS on the challenging nuScenes test set while maintaining high efficiency. For the first time, the NDS score of a camera model reaches 60%.