Camera Agnostic Two-Head Network for Ego-Lane Inference

TL;DR

This paper introduces a camera-agnostic, learning-based ego-lane inference method using a two-head network with attention mechanisms, enabling robust performance across diverse camera configurations without calibration.

Contribution

It presents a novel two-head network with attention guidance for ego-lane inference that does not depend on camera calibration, improving adaptability.

Findings

Validated across diverse environments and camera setups

Achieved robust ego-lane inference without calibration

Effective in varying camera orientations and mounting points

Abstract

Vision-based ego-lane inference using High-Definition (HD) maps is essential in autonomous driving and advanced driver assistance systems. The traditional approach necessitates well-calibrated cameras, which confines variation of camera configuration, as the algorithm relies on intrinsic and extrinsic calibration. In this paper, we propose a learning-based ego-lane inference by directly estimating the ego-lane index from a single image. To enhance robust performance, our model incorporates the two-head structure inferring ego-lane in two perspectives simultaneously. Furthermore, we utilize an attention mechanism guided by vanishing point-and-line to adapt to changes in viewpoint without requiring accurate calibration. The high adaptability of our model was validated in diverse environments, devices, and camera mounting points and orientations.