Faster Training, Fewer Labels: Self-Supervised Pretraining for Fine-Grained BEV Segmentation
更快的训练,更少的标签:用于细粒度BEV分割的自监督预训练
机构 * University of Wuppertal(乌珀塔尔大学) ; APTIV(APTIV公司)
专题命中 感知 :BEV(title,abstract);autonomous driving(abstract);分类 cs.CV
AI总结 本文提出一种自监督预训练方法,通过减少标注数据和训练时间,提升细粒度BEV分割性能。
Comments This Paper has been accepted to the 2026 IEEE Intelligent Vehicles Symposium (IV)