GeoBEV: Learning Geometric BEV Representation for Multi-view 3D Object Detection
This work addresses geometric limitations in BEV-based 3D object detection for autonomous driving applications, representing an incremental improvement over existing methods.
The paper tackles the problem of low geometric quality in Bird's-Eye-View (BEV) representations for multi-view 3D object detection by proposing Radial-Cartesian BEV Sampling, In-Box Labels, and Centroid-Aware Inner Loss, achieving a state-of-the-art result of 66.2% NDS on the nuScenes test set.
Bird's-Eye-View (BEV) representation has emerged as a mainstream paradigm for multi-view 3D object detection, demonstrating impressive perceptual capabilities. However, existing methods overlook the geometric quality of BEV representation, leaving it in a low-resolution state and failing to restore the authentic geometric information of the scene. In this paper, we identify the drawbacks of previous approaches that limit the geometric quality of BEV representation and propose Radial-Cartesian BEV Sampling (RC-Sampling), which outperforms other feature transformation methods in efficiently generating high-resolution dense BEV representation to restore fine-grained geometric information. Additionally, we design a novel In-Box Label to substitute the traditional depth label generated from the LiDAR points. This label reflects the actual geometric structure of objects rather than just their surfaces, injecting real-world geometric information into the BEV representation. In conjunction with the In-Box Label, Centroid-Aware Inner Loss (CAI Loss) is developed to capture the inner geometric structure of objects. Finally, we integrate the aforementioned modules into a novel multi-view 3D object detector, dubbed GeoBEV, which achieves a state-of-the-art result of 66.2\% NDS on the nuScenes test set. The code is available at https://github.com/mengtan00/GeoBEV.git.