An online HD map model that swaps ResNet and Swin backbones for DINOv2 and adds auxiliary BEV segmentation heads reports mAP gains of about 1.5 points over MapQR on nuScenes.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
MapFM: Foundation Model-Driven HD Mapping with Multi-Task Contextual Learning
An online HD map model that swaps ResNet and Swin backbones for DINOv2 and adds auxiliary BEV segmentation heads reports mAP gains of about 1.5 points over MapQR on nuScenes.