notesum.ai

Published at November 22

LiDAR-based End-to-end Temporal Perception for Vehicle-Infrastructure Cooperation

cs.CV

cs.AI

cs.RO

Released Date: November 22, 2024

Authors: Zhenwei Yang, Jilei Mao, Wenxian Yang, Yibo Ai, Yu Kong, Haibao Yu, Weidong Zhang

Arxiv: http://arxiv.org/abs/2411.14927v1

Refer to caption

Model	Latency 0ms			Latency 100ms			Latency 200ms			Latency 300ms
	mAP $\uparrow$	AMOTA $\uparrow$	AMOTP $\downarrow$	mAP $\uparrow$	AMOTA $\uparrow$	AMOTP $\downarrow$	mAP $\uparrow$	AMOTA $\uparrow$	AMOTP $\downarrow$	mAP $\uparrow$	AMOTA $\uparrow$	AMOTP $\downarrow$
LET-V (Ours)	0.456	0.467	1.113	-	-	-	-	-	-	-	-	-
V2VNet [18]	0.349	0.377	1.258	0.346	0.386	1.268	0.341	0.386	1.268	0.333	0.362	1.309
FFNet [22]	0.349	0.377	1.258	0.340	0.377	1.261	0.345	0.383	1.262	0.342	0.387	1.264
PointPillars [38]	0.469	0.509	0.822	0.456	0.510	0.833	0.444	0.486	0.875	0.441	0.478	0.874
LET-VIC (Ours)	0.606	0.640	0.700	0.579	0.631	0.716	0.554	0.613	0.775	0.526	0.570	0.811