Cross-View Supervision transfers geometric and topological priors from ego-aligned overhead perspectives into camera-based BEV encoders via feature-space alignment, yielding up to 44% relative mAP gains at long range on nuScenes.
In: Proceedings of the Neural Information Processing Systems Track on Datasets and Benchmarks (NeurIPS Datasets and Benchmarks 2021) (2021)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
dataset 1
citation-polarity summary
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1roles
dataset 1polarities
background 1representative citing papers
citing papers explorer
-
Learning Ego-Centric BEV Representations from a Perspective-Privileged View: Cross-View Supervision for Online HD Map Construction
Cross-View Supervision transfers geometric and topological priors from ego-aligned overhead perspectives into camera-based BEV encoders via feature-space alignment, yielding up to 44% relative mAP gains at long range on nuScenes.