Hierarchical and Decoupled BEV Perception Learning Framework for Autonomous Driving
This work addresses efficiency and customization issues in perception algorithm development for autonomous driving, though it appears incremental as it builds on existing BEV and deep learning approaches.
The paper tackles the challenges of lengthy development cycles, poor reusability, and complex sensor setups in autonomous driving perception by proposing a hierarchical BEV perception paradigm with a library of modules and graphical interface, resulting in significant improvement over traditional training schemes as demonstrated on the Nuscenes dataset.
Perception is essential for autonomous driving system. Recent approaches based on Bird's-eye-view (BEV) and deep learning have made significant progress. However, there exists challenging issues including lengthy development cycles, poor reusability, and complex sensor setups in perception algorithm development process. To tackle the above challenges, this paper proposes a novel hierarchical BEV perception paradigm, aiming to provide a library of fundamental perception modules and user-friendly graphical interface, enabling swift construction of customized models. We conduct the Pretrain-Finetune strategy to effectively utilize large scale public datasets and streamline development processes. Moreover, we present a Multi-Module Learning (MML) approach, enhancing performance through synergistic and iterative training of multiple models. Extensive experimental results on the Nuscenes dataset demonstrate that our approach renders significant improvement over the traditional training scheme.