notesum.ai

Published at December 6

EvTTC: An Event Camera Dataset for Time-to-Collision Estimation

cs.RO

cs.CV

Released Date: December 6, 2024

Authors: Kaizhen Sun¹, Jinghang Li¹, Kuan Dai¹, Bangyan Liao², Wei Xiong¹, Yi Zhou¹

Aff.: ¹Hunan University; ²Westlake University

Arxiv: http://arxiv.org/pdf/2412.05053v1

Refer to caption

Dataset	Frame Camera		Event Camera		Emergency	Groundtruth
Dataset	Resolution [pix]	Detection Range [m]	Resolution [pix]	Detection Range [m]	Break	Groundtruth
DR (eye) VE [1]	1920 $\times$ 1080	–	–	–	✖	Driver Attention Map
DADA-2000 [2]	1584 $\times$ 660	–	–	–	✔	Driver Attention Map
Crash to Not Crash [3]	710 $\times$ 400	–	–	–	✔	2D Bounding Box
TSTTC [4]	1024 $\times$ 576	[16-96]	–	–	✖	TTC
MVSEC [5]	752 $\times$ 480	28	346 $\times$ 260	12	✖	Depth, GPS
DSEC [6]	1440 $\times$ 1080	80	640 $\times$ 480	35	✖	Depth, RTK GPS
ViViD++ [7]	1280 $\times$ 1024	53	640 $\times$ 480	50	✖	Depth, RTK GPS
M3ED [8]	1280 $\times$ 800	68	1280 $\times$ 720	65	✖	Depth, RTK GPS
FCWD (Ours)	1920 $\times$ 1200	[160-295]	1280 $\times$ 720	[99-197]	✔	TTC, Depth, RTK GPS