The KITTI Vision Benchmark Suite

Time varies depending on the number of model components
Cars: 16 viewpoints ~ 28 seconds
Pedestrians: 8 viewpoints ~ 13 seconds
Cyclists: 4 viewpoints ~ 7 seconds

C=0.001

@article{Yebes2015SENSORS,
author = {J. Javier Yebes and Luis M. Bergasa and M.A. García-Garrido},
title = {{Visual Object Recognition with 3D-Aware Features in KITTI Urban Scenes}},
journal = {Sensors},
volume = {15},
number = {4},
pages = {9228--9250},
issn = {1424-8220},
year = {2015}
}

@inproceedings{ref:Yebes14iv,
author = {J.Javier Yebes and Luis M. Bergasa and Roberto Arroyo and Alberto Lázaro},
title = {{Supervised learning and evaluation of KITTI's cars detector with DPM}},
booktitle = IV,
pages = {768-773},
isbn = {978-1-4799-3637-3},
address = {Detroit, USA},
month = {June},
year = {2014}
}