Uncertainty-augmented LiDAR detection plus z-test on observation windows reduces false motion predictions versus velocity thresholding, with gains shown in real-world drives but parity on nuScenes.
Training independent subnetworks for robust prediction
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.RO 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Taming Perception Jitter: Uncertainty-Aware LiDAR Object Detection for Reliable Motion Classification
Uncertainty-augmented LiDAR detection plus z-test on observation windows reduces false motion predictions versus velocity thresholding, with gains shown in real-world drives but parity on nuScenes.