Camera Agnostic Two-Head Network for Ego-Lane Inference

arXiv:2404.12770v1 Announce Type: cross
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 …

abstract advanced advanced driver assistance algorithm arxiv autonomous autonomous driving cameras cs.cv cs.lg cs.ro definition driver driving head inference intrinsic maps network paper systems the algorithm type variation vision