中文GFSR：基于几何保真与空间细化的可靠车道线检测

车道检测研究中，现有方法仅依赖分类置信度过滤会导致几何质量差的候选被保留，而优化低置信度但高几何质量的候选被剔除。该工作揭示了这一问题并提出改进策略，通过结合几何质量重新定义置信度，提升复杂场景下车道检测鲁棒性，对自动驾驶安全性有重要实际意义。

arXiv:2605.23327v1 Announce Type: new Abstract: Lane detection stands as a crucial perception task in autonomous driving and advanced driver assistance systems. However, existing methods still degrade in complex real scenarios due to two major limitations. First, classification confidence only characterizes the categorical existence of lane candidates and has no strong correlation with geometric quality. If threshold filtering and NMS are conducted merely based on this confidence, the model tends to retain lane priors with high confidence while eliminating those with lower confidence but super