QuadBEV: An Efficient Quadruple-Task Perception Framework via Bird's-Eye-View Representation

TL;DR

QuadBEV is an efficient multitask perception framework for autonomous driving that integrates four key tasks into a shared system, reducing computational load and improving real-world applicability.

Contribution

It introduces a shared backbone architecture for four perception tasks, addressing multitask learning challenges and enhancing efficiency for resource-constrained environments.

Findings

Reduces redundant computations in perception tasks

Demonstrates robustness and effectiveness in experiments

Suitable for embedded autonomous driving systems

Abstract

Bird's-Eye-View (BEV) perception has become a vital component of autonomous driving systems due to its ability to integrate multiple sensor inputs into a unified representation, enhancing performance in various downstream tasks. However, the computational demands of BEV models pose challenges for real-world deployment in vehicles with limited resources. To address these limitations, we propose QuadBEV, an efficient multitask perception framework that leverages the shared spatial and contextual information across four key tasks: 3D object detection, lane detection, map segmentation, and occupancy prediction. QuadBEV not only streamlines the integration of these tasks using a shared backbone and task-specific heads but also addresses common multitask learning challenges such as learning rate sensitivity and conflicting task objectives. Our framework reduces redundant computations, thereby enhancing system efficiency, making it particularly suited for embedded systems. We present comprehensive experiments that validate the effectiveness and robustness of QuadBEV, demonstrating its suitability for real-world applications.