Structured Bird's-Eye-View Traffic Scene Understanding from Onboard Images

TL;DR

This paper introduces a method to extract a structured bird's-eye-view representation of road networks and traffic agents from a single onboard camera image, enhancing scene understanding for autonomous navigation.

Contribution

The work presents a novel approach to derive a directed graph of the road network and detect dynamic objects in BEV from monocular images, which is challenging due to camera mounting orientation.

Findings

Achieves superior performance over baselines

Effectively detects road graph and objects in BEV

Demonstrates robustness through ablation studies

Abstract

Autonomous navigation requires structured representation of the road network and instance-wise identification of the other traffic agents. Since the traffic scene is defined on the ground plane, this corresponds to scene understanding in the bird's-eye-view (BEV). However, the onboard cameras of autonomous cars are customarily mounted horizontally for a better view of the surrounding, making this task very challenging. In this work, we study the problem of extracting a directed graph representing the local road network in BEV coordinates, from a single onboard camera image. Moreover, we show that the method can be extended to detect dynamic objects on the BEV plane. The semantics, locations, and orientations of the detected objects together with the road graph facilitates a comprehensive understanding of the scene. Such understanding becomes fundamental for the downstream tasks, such as path planning and navigation. We validate our approach against powerful baselines and show that our network achieves superior performance. We also demonstrate the effects of various design choices through ablation studies. Code: https://github.com/ybarancan/STSU