A BEVFusion-based offboard tracker with density-aware loss weighting, nearest-neighbor relationship targets, and high-resolution sparse features doubles MOTA on a new crowded-pedestrian benchmark (PCP-MV) from 0.172 to 0.353.
An end-to-end transformer model for 3d object detection,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Learning better representations for crowded pedestrians in offboard LiDAR-camera 3D tracking-by-detection
A BEVFusion-based offboard tracker with density-aware loss weighting, nearest-neighbor relationship targets, and high-resolution sparse features doubles MOTA on a new crowded-pedestrian benchmark (PCP-MV) from 0.172 to 0.353.