The KITTI Vision Benchmark Suite

The object detection and object orientation estimation benchmark consists of 7481 training images and 7518 test images, comprising a total of 80.256 labeled objects. All images are color and saved as png. For evaluation, we compute precision-recall curves for object detection and orientation-similarity-recall curves for joint object detection and orientation estimation. In the latter case not only the object 2D bounding box has to be located correctly, but also the orientation estimate in bird's eye view is evaluated. To rank the methods we compute average precision and average orientation similiarity. We require that all methods use the same parameter set for all test pairs. Our development kit provides details about the data format as well as MATLAB / C++ utility functions for reading and writing the label files.

• Download left color images of object data set (12 GB)
• Download right color images, if you want to use stereo information (12 GB)
• Download the 3 temporally preceding frames (left color) (36 GB)
• Download the 3 temporally preceding frames (right color) (36 GB)
• Download Velodyne point clouds, if you want to use laser information (29 GB)
• Download camera calibration matrices of object data set (16 MB)
• Download training labels of object data set (5 MB)
• Download object development kit (1 MB)
• Download pre-trained LSVM baseline models (5 MB) used in Joint 3D Estimation of Objects and Scene Layout (NIPS 2011). These models are referred to as LSVM-MDPM-sv (supervised version) and LSVM-MDPM-us (unsupervised version) in the tables below.
• Download reference detections (L-SVM) for training and test set (800 MB)
• Qianli Liao (NYU) has put together code to convert from KITTI to PASCAL VOC file format (documentation included, requires Emacs).
• Karl Rosaen (U.Mich) has released code to convert between KITTI, KITTI tracking, Pascal VOC, Udacity, CrowdAI and AUTTI formats.

We evaluate object detection performance using the PASCAL criteria and object detection and orientation estimation performance using the measure discussed in our CVPR 2012 publication. For cars we require an overlap of 70%, while for pedestrians and cyclists we require an overlap of 50% for a detection. Detections in don't care areas or detections which are smaller than the minimum size do not count as false positive. Difficulties are defined as follows:

• Easy: Min. bounding box height: 40 Px, Max. occlusion level: Fully visible, Max. truncation: 15 %
• Moderate: Min. bounding box height: 25 Px, Max. occlusion level: Partly occluded, Max. truncation: 30 %
• Hard: Min. bounding box height: 25 Px, Max. occlusion level: Difficult to see, Max. truncation: 50 %

All methods are ranked based on the moderately difficult results. Note that for the hard evaluation ~2 % of the provided bounding boxes have not been recognized by humans, thereby upper bounding recall at 98 %. Hence, the hard evaluation is only given for reference.

Car

	Method	Setting	Code	Moderate	Easy	Hard	Runtime	Environment
1	DuEye			92.65 %	91.43 %	86.18 %	4 s	GPU @ 2.5 Ghz (C/C++)
2	BM-NET			91.84 %	91.36 %	81.98 %	0.5 s	GPU @ 2.5 Ghz (C/C++)
3	RV-CNN			91.67 %	91.28 %	85.43 %	3.5 s	GPU @ 2.5 Ghz (Python + C/C++)
4	eagle			91.28 %	91.06 %	85.66 %	4 s	GPU @ 2.5 Ghz (C/C++)
5	Genome			90.63 %	90.85 %	85.82 %	4 s	GPU @ 2.5 Ghz (C/C++)
6	TuSimple			90.34 %	91.30 %	82.63 %	1.6 s	GPU @ 2.5 Ghz (Python + C/C++)
7	SAIT			90.32 %	92.97 %	80.39 %	0.15 s	GPU @ >3.5 Ghz (Python + C/C++)
8	RRC		code	90.19 %	93.66 %	86.97 %	3.6 s	GPU @ 2.5 Ghz (Python + C/C++)
J. Ren, X. Chen, J. Liu, W. Sun, J. Pang, Q. Yan, Y. Tai and L. Xu: Accurate Single Stage Detector Using Recurrent Rolling Convolution. CVPR 2017.
9	Direwolf			90.06 %	96.41 %	80.64 %	0.5 s	GPU @ 2.5 Ghz (C/C++)
10	Deep MANTA			90.03 %	95.77 %	80.62 %	0.7 s	GPU @ 2.5 Ghz (Python + C/C++)
F. Chabot, M. Chaouch, J. Rabarisoa, C. Teulière and T. Chateau: Deep MANTA: A Coarse-to-fine Many-Task Network for joint 2D and 3D vehicle analysis from monocular image. CVPR 2017.
11	sensekitti			90.03 %	91.19 %	81.69 %	4.5 s	GPU @ 2.5 Ghz (Python + C/C++)
12	NVDriveNet-H			89.81 %	90.92 %	83.76 %	0.15s	GPU @ 2.5 Ghz (Python + C/C++)
13	Allspark			89.72 %	90.36 %	78.89 %	0.7 s	GPU @ 2.5 Ghz (C/C++)
14	SINet_VGG			89.60 %	90.60 %	77.75 %	0.2 s	GPU @ 2.5 Ghz (Matlab + C/C++)
15	DJML			89.39 %	91.18 %	79.59 %	2.4 s	GPU @ 2.5 Ghz (Python + C/C++)
16	HSR2			89.34 %	94.46 %	79.10 %	0.15 s	1 core @ 2.5 Ghz (C/C++)
17	uickitti			89.23 %	90.83 %	79.46 %	1.5 s	GPU @ 2.5 Ghz (C/C++)
18	SINet_PVA			89.21 %	91.91 %	76.33 %	0.11 s	GPU @ 2.5 Ghz (Matlab + C/C++)
19	Pie			89.19 %	89.39 %	78.72 %	1.2 s	1 core @ 2.5 Ghz (C/C++)
20	Deep3DBox			89.04 %	92.98 %	77.17 %	1.5 s	GPU @ 2.5 Ghz (C/C++)
A. Mousavian, D. Anguelov, J. Flynn and J. Kosecka: 3D Bounding Box Estimation Using Deep Learning and Geometry. CVPR 2017.
21	DeepStereoOP			89.04 %	93.45 %	79.58 %	3.4 s	GPU @ 3.5 Ghz (Matlab + C/C++)
C. Pham and J. Jeon: Robust Object Proposals Re-ranking for Object Detection in Autonomous Driving Using Convolutional Neural Networks. Signal Processing: Image Communiation 2017.
22	SubCNN			89.04 %	90.81 %	79.27 %	2 s	GPU @ 3.5 Ghz (Python + C/C++)
Y. Xiang, W. Choi, Y. Lin and S. Savarese: Subcategory-aware Convolutional Neural Networks for Object Proposals and Detection. IEEE Winter Conference on Applications of Computer Vision (WACV) 2017.
23	MS-CNN		code	89.02 %	90.03 %	76.11 %	0.4 s	GPU @ 2.5 Ghz (C/C++)
Z. Cai, Q. Fan, R. Feris and N. Vasconcelos: A Unified Multi-scale Deep Convolutional Neural Network for Fast Object Detection. ECCV 2016.
24	HM_SSD_RCNN			89.02 %	94.41 %	78.84 %	0.15 s	1 core @ 2.5 Ghz (C/C++)
25	CPCD			88.88 %	93.48 %	79.33 %	3 s	1 core @ 2.5 Ghz (C/C++)
26	SDP+RPN			88.85 %	90.14 %	78.38 %	0.4 s	GPU @ 2.5 Ghz (Python + C/C++)
F. Yang, W. Choi and Y. Lin: Exploit All the Layers: Fast and Accurate CNN Object Detector with Scale Dependent Pooling and Cascaded Rejection Classifiers. Proceedings of the IEEE International Conference on Computer Vision and Pattern Recognition 2016. S. Ren, K. He, R. Girshick and J. Sun: Faster R-CNN: Towards real-time object detection with region proposal networks. Advances in Neural Information Processing Systems 2015.
27	Re-3DOP			88.72 %	93.39 %	79.22 %	3 s	1 core @ 2.5 Ghz (C/C++)
28	Mono3D		code	88.66 %	92.33 %	78.96 %	4.2 s	GPU @ 2.5 Ghz (Matlab + C/C++)
X. Chen, K. Kundu, Z. Zhang, H. Ma, S. Fidler and R. Urtasun: Monocular 3D Object Detection for Autonomous Driving. CVPR 2016.
29	3DOP		code	88.64 %	93.04 %	79.10 %	3s	GPU @ 2.5 Ghz (Matlab + C/C++)
X. Chen, K. Kundu, Y. Zhu, A. Berneshawi, H. Ma, S. Fidler and R. Urtasun: 3D Object Proposals for Accurate Object Class Detection. NIPS 2015.
30	MM-MRFC			88.45 %	90.63 %	78.32 %	0.05 s	GPU @ 2.5 Ghz (C/C++)
31	SYVO			88.34 %	89.44 %	72.04 %	0.13 s	GPU @ 2.5 Ghz (C/C++)
32	MV3D			87.67 %	89.11 %	79.54 %	0.36 s	GPU @ 2.5 Ghz (Python + C/C++)
X. Chen, H. Ma, J. Wan, B. Li and T. Xia: Multi-View 3D Object Detection Network for Autonomous Driving. CVPR 2017.
33	TWSNet			87.61 %	89.19 %	71.19 %	0.48 s	GPU @ 3.5 Ghz (Matlab + C/C++)
34	WRInception			87.46 %	88.57 %	77.79 %	0.06 s	GPU @ 2.5 Ghz (C/C++)
35	CNN based			87.25 %	87.75 %	77.38 %	1s	1 core @ 2.5 Ghz (C/C++)
36	UI			87.18 %	87.83 %	78.35 %	0.4 s	GPU @ 2.5 Ghz (C/C++)
37	tbd			87.13 %	91.85 %	78.16 %	1 s	1 core @ 2.5 Ghz (C/C++)
38	ANM			85.61 %	86.80 %	77.20 %	0.05 s	GPU @ 2.5 Ghz (C/C++)
39	PNET			83.60 %	81.77 %	74.07 %	0.1 s	GPU @ 2.5 Ghz (Python + C/C++)
40	SDP+CRC (ft)			83.53 %	90.33 %	71.13 %	0.6 s	GPU @ 2.5 Ghz (C/C++)
F. Yang, W. Choi and Y. Lin: Exploit All the Layers: Fast and Accurate CNN Object Detector with Scale Dependent Pooling and Cascaded Rejection Classifiers. Proceedings of the IEEE International Conference on Computer Vision and Pattern Recognition 2016.
41	Faster R-CNN		code	81.84 %	86.71 %	71.12 %	2 s	GPU @ 3.5 Ghz (Python + C/C++)
S. Ren, K. He, R. Girshick and J. Sun: Faster R-CNN: Towards Real- Time Object Detection with Region Proposal Networks. NIPS 2015.
42	MV3D (LIDAR)			79.24 %	87.00 %	78.16 %	0.24 s	GPU @ 2.5 Ghz (Python + C/C++)
X. Chen, H. Ma, J. Wan, B. Li and T. Xia: Multi-View 3D Object Detection Network for Autonomous Driving. CVPR 2017.
43	RefineNet			79.17 %	89.88 %	66.38 %	0.20 s	GPU @ 2.5 Ghz (Matlab + C++)
R. Rajaram, E. Bar and M. Trivedi: RefineNet: Iterative Refinement for Accurate Object Localization. Intelligent Transportation Systems Conference 2016.
44	spLBP			77.39 %	87.18 %	60.59 %	1.5 s	8 cores @ 2.5 Ghz (Matlab + C/C++)
Q. Hu, S. Paisitkriangkrai, C. Shen, A. Hengel and F. Porikli: Fast Detection of Multiple Objects in Traffic Scenes With a Common Detection Framework. IEEE Trans. Intelligent Transportation Systems 2016.
45	SceneNet			77.22 %	86.91 %	68.24 %	0.03 s	GPU @ 2.5 Ghz (C/C++)
46	Reinspect		code	76.65 %	88.13 %	66.23 %	2s	1 core @ 2.5 Ghz (C/C++)
R. Stewart, M. Andriluka and A. Ng: End-to-End People Detection in Crowded Scenes. CVPR 2016.
47	Regionlets			76.45 %	84.75 %	59.70 %	1 s	>8 cores @ 2.5 Ghz (C/C++)
X. Wang, M. Yang, S. Zhu and Y. Lin: Regionlets for Generic Object Detection. T-PAMI 2015. W. Zou, X. Wang, M. Sun and Y. Lin: Generic Object Detection with Dense Neural Patterns and Regionlets. British Machine Vision Conference 2014. C. Long, X. Wang, G. Hua, M. Yang and Y. Lin: Accurate Object Detection with Location Relaxation and Regionlets Relocalization. Asian Conference on Computer Vision 2014.
48	AOG		code	75.94 %	84.80 %	60.70 %	3 s	4 cores @ 2.5 Ghz (Matlab)
T. Wu, B. Li and S. Zhu: Learning And-Or Models to Represent Context and Occlusion for Car Detection and Viewpoint Estimation. TPAMI 2016. B. Li, T. Wu and S. Zhu: Integrating Context and Occlusion for Car Detection by Hierarchical And-Or Model. ECCV 2014.
49	Pose-RCNN			75.80 %	88.43 %	66.57 %	2 s	>8 cores @ 2.5 Ghz (Python)
50	3DVP		code	75.77 %	87.46 %	65.38 %	40 s	8 cores @ 3.5 Ghz (Matlab + C/C++)
Y. Xiang, W. Choi, Y. Lin and S. Savarese: Data-Driven 3D Voxel Patterns for Object Category Recognition. IEEE Conference on Computer Vision and Pattern Recognition 2015.
51	SubCat		code	75.46 %	84.14 %	59.71 %	0.7 s	6 cores @ 3.5 Ghz (Matlab + C/C++)
E. Ohn-Bar and M. Trivedi: Learning to Detect Vehicles by Clustering Appearance Patterns. T-ITS 2015.
52	AR-FCN			75.43 %	79.69 %	65.75 %	0.19 s	GPU @ 2.5 Ghz (C/C++)
53	XXX			75.34 %	84.16 %	67.98 %	>5 s	1 core @ 2.5 Ghz (C/C++)
54	FD2			74.81 %	87.14 %	66.08 %	0.01 s	GPU @ >3.5 Ghz (Python + C/C++)
55	FD			71.59 %	78.05 %	62.01 %	0.01 s	GPU @ >3.5 Ghz (Python)
56	AOG-View			71.16 %	83.38 %	57.11 %	3 s	1 core @ 2.5 Ghz (Matlab, C/C++)
B. Li, T. Wu and S. Zhu: Integrating Context and Occlusion for Car Detection by Hierarchical And-Or Model. ECCV 2014.
57	FCNN			70.67 %	87.69 %	61.49 %	0.1 s	1 core @ 2.5 Ghz (C/C++)
58	MV-RGBD-RF			69.92 %	76.40 %	57.47 %	4 s	4 cores @ 2.5 Ghz (C/C++)
A. Gonzalez, D. Vazquez, A. Lopez and J. Amores: On-Board Object Detection: Multicue, Multimodal, and Multiview Random Forest of Local Experts.. IEEE Trans. on Cybernetics 2016. A. Gonzalez, G. Villalonga, J. Xu, D. Vazquez, J. Amores and A. Lopez: Multiview Random Forest of Local Experts Combining RGB and LIDAR data for Pedestrian Detection. IEEE Intelligent Vehicles Symposium (IV) 2015.
59	SmartCNN			69.54 %	67.29 %	55.88 %	1 s	1 core @ 2.5 Ghz (C/C++)
60	Vote3Deep			68.24 %	76.79 %	63.23 %	1.5 s	4 cores @ 2.5 Ghz (C/C++)
M. Engelcke, D. Rao, D. Zeng Wang, C. Hay Tong and I. Posner: Vote3Deep: Fast Object Detection in 3D Point Clouds Using Efficient Convolutional Neural Networks. ArXiv e-prints 2016.
61	GPVL			67.89 %	80.53 %	58.23 %	10 s	1 core @ 2.5 Ghz (C/C++)
62	GVPL			67.78 %	79.88 %	57.69 %	1 s	8 cores @ 2.5 Ghz (Matlab + C/C++)
63	ZGC			67.75 %	84.69 %	58.57 %	0.12 s	1 core @ 2.5 Ghz (C/C++)
64	QHY			67.55 %	84.53 %	58.28 %	0.1 s	1 core @ 2.5 Ghz (C/C++)
65	BdCost48LDCF			66.66 %	77.37 %	55.51 %	5 s	1 core @ 2.5 Ghz (C/C++)
66	BdCost48-25C			66.27 %	77.59 %	55.68 %	4 s	1 core @ 2.5 Ghz (C/C++)
67	OC-DPM			65.95 %	74.94 %	53.86 %	10 s	8 cores @ 2.5 Ghz (Matlab)
B. Pepik, M. Stark, P. Gehler and B. Schiele: Occlusion Patterns for Object Class Detection. IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2013.
68	RCNN			65.43 %	82.20 %	50.92 %	0.08 s	GPU @ 2.5 Ghz (Python + C/C++)
69	DPM-VOC+VP			64.71 %	74.95 %	48.76 %	8 s	1 core @ 2.5 Ghz (C/C++)
B. Pepik, M. Stark, P. Gehler and B. Schiele: Multi-view and 3D Deformable Part Models. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) 2015.
70	HL			64.69 %	80.59 %	50.50 %	0.16 s	1 core @ 2.5 Ghz (C/C++)
71	LCNN			64.53 %	78.72 %	56.52 %	1 s	1 core @ 2.5 Ghz (C/C++)
72	NMF-CNN			62.88 %	79.07 %	52.67 %	0.1 s	GPU @ 2.5 Ghz (Matlab + C/C++)
73	MDPM-un-BB			62.16 %	71.19 %	48.43 %	60 s	4 core @ 2.5 Ghz (MATLAB)
P. Felzenszwalb, R. Girshick, D. McAllester and D. Ramanan: Object Detection with Discriminatively Trained Part-Based Models. PAMI 2010.
74	SubCat48LDCF			61.79 %	68.71 %	47.46 %	5 s	1 core @ 2.5 Ghz (Matlab + C/C++)
75	NMRDO			61.72 %	72.42 %	54.06 %	0.1 s	GPU @ 2.5 Ghz (Python + C/C++)
76	DPM-C8B1			60.99 %	74.33 %	47.16 %	15 s	4 cores @ 2.5 Ghz (Matlab + C/C++)
J. Yebes, L. Bergasa and M. García-Garrido: Visual Object Recognition with 3D-Aware Features in KITTI Urban Scenes. Sensors 2015. J. Yebes, L. Bergasa, R. Arroyo and A. Lázaro: Supervised learning and evaluation of KITTI's cars detector with DPM. IV 2014.
77	HgCNN			60.10 %	65.83 %	52.29 %	1 s	1 core @ 2.5 Ghz (C/C++)
78	ACF-SC			58.66 %	69.11 %	45.95 %	<0.3 s	1 core @ >3.5 Ghz (Matlab + C/C++)
C. Cadena, A. Dick and I. Reid: A Fast, Modular Scene Understanding System using Context-Aware Object Detection. Robotics and Automation (ICRA), 2015 IEEE International Conference on 2015.
79	LSVM-MDPM-sv			57.44 %	71.70 %	46.58 %	10 s	4 cores @ 3.0 Ghz (C/C++)
P. Felzenszwalb, R. Girshick, D. McAllester and D. Ramanan: Object Detection with Discriminatively Trained Part-Based Models. PAMI 2010. A. Geiger, C. Wojek and R. Urtasun: Joint 3D Estimation of Objects and Scene Layout. NIPS 2011.
80	frd			55.80 %	68.87 %	48.04 %	2 s	1 core @ 2.5 Ghz (C/C++)
81	LSVM-MDPM-us		code	55.42 %	66.53 %	41.04 %	10 s	4 cores @ 3.0 Ghz (C/C++)
P. Felzenszwalb, R. Girshick, D. McAllester and D. Ramanan: Object Detection with Discriminatively Trained Part-Based Models. PAMI 2010.
82	ACF			54.74 %	55.89 %	42.98 %	0.2 s	1 core @ >3.5 Ghz (Matlab + C/C++)
P. Doll\'ar, R. Appel, S. Belongie and P. Perona: Fast Feature Pyramids for Object Detection. PAMI 2014. P. Doll\'ar: Piotr's Image and Video Matlab Toolbox (PMT). .
83	VeloFCN			53.59 %	71.06 %	46.92 %	1 s	GPU @ 2.5 Ghz (Python + C/C++)
B. Li, T. Zhang and T. Xia: Vehicle Detection from 3D Lidar Using Fully Convolutional Network. RSS 2016 .
84	MLSmoke			52.59 %	69.30 %	43.82 %	1 s	1 core @ 2.5 Ghz (C/C++)
85	Vote3D			47.99 %	56.80 %	42.57 %	0.5 s	4 cores @ 2.8 Ghz (C/C++)
D. Wang and I. Posner: Voting for Voting in Online Point Cloud Object Detection. Proceedings of Robotics: Science and Systems 2015.
86	YOLO			35.74 %	47.69 %	29.65 %	0.03 s	GPU @ 1.0 Ghz (C/C++)
87	CSoR		code	26.13 %	34.79 %	22.69 %	3.5 s	4 cores @ >3.5 Ghz (Python + C/C++)
L. Plotkin: PyDriver: Entwicklung eines Frameworks für räumliche Detektion und Klassifikation von Objekten in Fahrzeugumgebung. 2015.
88	R-CNN_VGG			26.06 %	32.40 %	20.91 %	10 s	GPU @ 2.5 Ghz (Matlab + C/C++)
89	mBoW			23.76 %	36.02 %	18.44 %	10 s	1 core @ 2.5 Ghz (C/C++)
J. Behley, V. Steinhage and A. Cremers: Laser-based Segment Classification Using a Mixture of Bag-of-Words. Proc. of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2013.
90	YOLOv2		code	19.21 %	28.35 %	15.94 %	0.02 s	GPU @ 3.5 Ghz (C/C++)
J. Redmon, S. Divvala, R. Girshick and A. Farhadi: You only look once: Unified, real-time object detection. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2016. J. Redmon and A. Farhadi: YOLO9000: Better, Faster, Stronger. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2017.

Pedestrian

	Method	Setting	Code	Moderate	Easy	Hard	Runtime	Environment
1	TuSimple			77.03 %	86.65 %	72.42 %	1.6 s	GPU @ 2.5 Ghz (Python + C/C++)
2	RRC		code	75.33 %	84.95 %	70.39 %	3.6 s	GPU @ 2.5 Ghz (Python + C/C++)
J. Ren, X. Chen, J. Liu, W. Sun, J. Pang, Q. Yan, Y. Tai and L. Xu: Accurate Single Stage Detector Using Recurrent Rolling Convolution. CVPR 2017.
3	iFDT			74.30 %	85.04 %	68.86 %	2.4 s	GPU @ 2.5 Ghz (Python + C/C++)
4	Allspark			74.24 %	84.73 %	68.60 %	0.7 s	GPU @ 2.5 Ghz (C/C++)
5	TiCNN			74.02 %	83.97 %	68.87 %	0.5 s	GPU @ 2.5 Ghz (Matlab + C/C++)
6	MS-CNN		code	73.70 %	83.92 %	68.31 %	0.4 s	GPU @ 2.5 Ghz (C/C++)
Z. Cai, Q. Fan, R. Feris and N. Vasconcelos: A Unified Multi-scale Deep Convolutional Neural Network for Fast Object Detection. ECCV 2016.
7	Pie			73.17 %	84.86 %	67.55 %	1.2 s	1 core @ 2.5 Ghz (C/C++)
8	SAIT			72.62 %	84.54 %	67.94 %	0.15 s	GPU @ >3.5 Ghz (Python + C/C++)
9	uickitti			71.84 %	83.49 %	67.00 %	1.5 s	GPU @ 2.5 Ghz (C/C++)
10	GN			71.65 %	82.03 %	65.00 %	1 s	GPU @ 2.5 Ghz (Matlab + C/C++)
11	SubCNN			71.33 %	83.28 %	66.36 %	2 s	GPU @ 3.5 Ghz (Python + C/C++)
Y. Xiang, W. Choi, Y. Lin and S. Savarese: Subcategory-aware Convolutional Neural Networks for Object Proposals and Detection. IEEE Winter Conference on Applications of Computer Vision (WACV) 2017.
12	IVA		code	70.70 %	83.63 %	64.67 %	0.4 s	GPU @ 2.5 Ghz (C/C++)
Y. Zhu, J. Wang, C. Zhao, H. Guo and H. Lu: Scale-adaptive Deconvolutional Regression Network for Pedestrian Detection. ACCV 2016. S. Ren, K. He, R. Girshick and J. Sun: Faster R-CNN: Towards real-time object detection with region proposal networks. Advances in neural information processing systems 2015.
13	SDP+RPN			70.16 %	80.09 %	64.82 %	0.4 s	GPU @ 2.5 Ghz (Python + C/C++)
F. Yang, W. Choi and Y. Lin: Exploit All the Layers: Fast and Accurate CNN Object Detector with Scale Dependent Pooling and Cascaded Rejection Classifiers. Proceedings of the IEEE International Conference on Computer Vision and Pattern Recognition 2016. S. Ren, K. He, R. Girshick and J. Sun: Faster R-CNN: Towards real-time object detection with region proposal networks. Advances in Neural Information Processing Systems 2015.
14	MM-MRFC			70.02 %	82.18 %	64.74 %	0.05 s	GPU @ 2.5 Ghz (C/C++)
15	WRInception			68.72 %	79.94 %	63.44 %	0.06 s	GPU @ 2.5 Ghz (C/C++)
16	3DOP		code	67.47 %	81.78 %	64.70 %	3s	GPU @ 2.5 Ghz (Matlab + C/C++)
X. Chen, K. Kundu, Y. Zhu, A. Berneshawi, H. Ma, S. Fidler and R. Urtasun: 3D Object Proposals for Accurate Object Class Detection. NIPS 2015.
17	DeepStereoOP			67.32 %	81.82 %	65.12 %	3.4 s	GPU @ 3.5 Ghz (Matlab + C/C++)
C. Pham and J. Jeon: Robust Object Proposals Re-ranking for Object Detection in Autonomous Driving Using Convolutional Neural Networks. Signal Processing: Image Communiation 2017.
18	sensekitti			67.29 %	79.58 %	62.28 %	4.5 s	GPU @ 2.5 Ghz (Python + C/C++)
19	Re-3DOP			67.27 %	80.87 %	64.02 %	3 s	1 core @ 2.5 Ghz (C/C++)
20	Mono3D		code	66.68 %	80.35 %	63.44 %	4.2 s	GPU @ 2.5 Ghz (Matlab + C/C++)
X. Chen, K. Kundu, Z. Zhang, H. Ma, S. Fidler and R. Urtasun: Monocular 3D Object Detection for Autonomous Driving. CVPR 2016.
21	IVA		code	66.50 %	78.09 %	61.60 %	1 s	GPU @ 2.5 Ghz (Matlab + C/C++)
22	HM_SSD_RCNN			66.40 %	81.92 %	59.21 %	0.15 s	1 core @ 2.5 Ghz (C/C++)
23	HSR2			65.91 %	81.02 %	63.03 %	0.15 s	1 core @ 2.5 Ghz (C/C++)
24	Faster R-CNN		code	65.90 %	78.86 %	61.18 %	2 s	GPU @ 3.5 Ghz (Python + C/C++)
S. Ren, K. He, R. Girshick and J. Sun: Faster R-CNN: Towards Real- Time Object Detection with Region Proposal Networks. NIPS 2015.
25	Tx			65.06 %	77.33 %	59.48 %	2 s	GPU @ 2.5 Ghz (Matlab + C/C++)
26	DJML			64.91 %	76.56 %	58.96 %	2.4 s	GPU @ 2.5 Ghz (Python + C/C++)
27	PNET			64.66 %	77.16 %	60.40 %	0.1 s	GPU @ 2.5 Ghz (Python + C/C++)
28	tbd			64.56 %	79.58 %	61.27 %	1 s	1 core @ 2.5 Ghz (C/C++)
29	SDP+CRC (ft)			64.19 %	77.74 %	59.27 %	0.6 s	GPU @ 2.5 Ghz (C/C++)
F. Yang, W. Choi and Y. Lin: Exploit All the Layers: Fast and Accurate CNN Object Detector with Scale Dependent Pooling and Cascaded Rejection Classifiers. Proceedings of the IEEE International Conference on Computer Vision and Pattern Recognition 2016.
30	Pose-RCNN			63.40 %	77.53 %	57.49 %	2 s	>8 cores @ 2.5 Ghz (Python)
31	CFM			63.26 %	74.22 %	56.44 %	<2 s	GPU @ 2.5 Ghz (Matlab + C/C++)
Q. Hu, P. Wang, C. Shen, A. Hengel and F. Porikli: Pushing the Limits of Deep CNNs for Pedestrian Detection. IEEE Transactions on Circuits and Systems for Video Technology 2017.
32	PCN			62.08 %	74.56 %	56.68 %	0.6 s
33	RPN+BF		code	61.29 %	75.45 %	56.08 %	0.6 s	GPU @ 2.5 Ghz (Matlab + C/C++)
L. Zhang, L. Lin, X. Liang and K. He: Is Faster R-CNN Doing Well for Pedestrian Detection?. ECCV 2016.
34	RB			61.15 %	77.12 %	55.12 %	0.6 s	GPU @ 2.5 Ghz (Matlab + C/C++)
35	Regionlets			61.15 %	73.14 %	55.21 %	1 s	>8 cores @ 2.5 Ghz (C/C++)
X. Wang, M. Yang, S. Zhu and Y. Lin: Regionlets for Generic Object Detection. T-PAMI 2015. W. Zou, X. Wang, M. Sun and Y. Lin: Generic Object Detection with Dense Neural Patterns and Regionlets. British Machine Vision Conference 2014. C. Long, X. Wang, G. Hua, M. Yang and Y. Lin: Accurate Object Detection with Location Relaxation and Regionlets Relocalization. Asian Conference on Computer Vision 2014.
36	LC			60.67 %	69.89 %	54.47 %	1 s	1 core @ 2.5 Ghz (Matlab + C/C++)
37	ens			60.64 %	72.14 %	54.59 %
38	CompACT-Deep			58.74 %	70.69 %	52.71 %	1 s	1 core @ 2.5 Ghz (Matlab + C/C++)
Z. Cai, M. Saberian and N. Vasconcelos: Learning Complexity-Aware Cascades for Deep Pedestrian Detection. ICCV 2015.
39	FichaDet			58.69 %	69.50 %	52.97 %	0.2 s	4 cores @ 2.5 Ghz (C/C++)
40	DeepParts			58.67 %	70.49 %	52.78 %	~1 s	GPU @ 2.5 Ghz (Matlab)
Y. Tian, P. Luo, X. Wang and X. Tang: Deep Learning Strong Parts for Pedestrian Detection. ICCV 2015.
41	p2dv			56.98 %	68.48 %	50.99 %	1 s	1 core @ 2.5 Ghz (C/C++)
42	D-TSF			56.77 %	68.44 %	50.77 %	1 s	1 core @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
43	FilteredICF			56.75 %	67.65 %	51.12 %	~ 2 s	>8 cores @ 2.5 Ghz (Matlab + C/C++)
S. Zhang, R. Benenson and B. Schiele: Filtered Channel Features for Pedestrian Detection. CVPR 2015.
44	FD2			56.65 %	71.11 %	51.62 %	0.01 s	GPU @ >3.5 Ghz (Python + C/C++)
45	MV-RGBD-RF			56.59 %	73.30 %	49.63 %	4 s	4 cores @ 2.5 Ghz (C/C++)
A. Gonzalez, D. Vazquez, A. Lopez and J. Amores: On-Board Object Detection: Multicue, Multimodal, and Multiview Random Forest of Local Experts.. IEEE Trans. on Cybernetics 2016. A. Gonzalez, G. Villalonga, J. Xu, D. Vazquez, J. Amores and A. Lopez: Multiview Random Forest of Local Experts Combining RGB and LIDAR data for Pedestrian Detection. IEEE Intelligent Vehicles Symposium (IV) 2015.
46	ACNet+Cascad			56.23 %	64.80 %	50.67 %	2.5 s	1 core @ 3.5 Ghz (Matlab)
47	Vote3Deep			55.37 %	68.39 %	52.59 %	1.5 s	4 cores @ 2.5 Ghz (C/C++)
M. Engelcke, D. Rao, D. Zeng Wang, C. Hay Tong and I. Posner: Vote3Deep: Fast Object Detection in 3D Point Clouds Using Efficient Convolutional Neural Networks. ArXiv e-prints 2016.
48	FD			55.10 %	66.84 %	49.82 %	0.01 s	GPU @ >3.5 Ghz (Python)
49	pAUCEnsT			54.49 %	65.26 %	48.60 %	60 s	1 core @ 2.5 Ghz (Matlab + C/C++)
S. Paisitkriangkrai, C. Shen and A. Hengel: Pedestrian Detection with Spatially Pooled Features and Structured Ensemble Learning. arXiv 2014.
50	PDV2			53.74 %	65.39 %	49.47 %	3.7 s	1 core @ 3.0 Ghz Matlab (C/C++)
J. Shen, X. Zuo, J. Li, W. Yang and H. Ling: A novel pixel neighborhood differential statistic feature for pedestrian and face detection . Pattern Recognition 2017.
51	ANM			52.55 %	69.86 %	51.13 %	0.05 s	GPU @ 2.5 Ghz (C/C++)
52	ACFD		code	50.91 %	61.61 %	45.51 %	0.2 s	4 cores @ >3.5 Ghz (C/C++)
53	ZGC			50.42 %	66.84 %	42.79 %	0.12 s	1 core @ 2.5 Ghz (C/C++)
54	R-CNN			50.13 %	61.61 %	44.79 %	4 s	GPU @ 3.3 Ghz (C/C++)
J. Hosang, M. Omran, R. Benenson and B. Schiele: Taking a Deeper Look at Pedestrians. arXiv 2015.
55	SSD1			50.03 %	63.78 %	47.15 %	0.255 s	GPU @ 2.5 Ghz (python+ C/C++)
56	NMF-CNN			49.26 %	65.16 %	45.51 %	0.1 s	GPU @ 2.5 Ghz (Matlab + C/C++)
57	ACF			47.29 %	60.11 %	42.90 %	1 s	1 core @ 3.5 Ghz (Matlab + C/C++)
P. Doll\'ar, R. Appel, S. Belongie and P. Perona: Fast Feature Pyramids for Object Detection. PAMI 2014.
58	Fusion-DPM		code	46.67 %	59.51 %	42.05 %	~ 30 s	1 core @ 3.5 Ghz (Matlab + C/C++)
C. Premebida, J. Carreira, J. Batista and U. Nunes: Pedestrian Detection Combining RGB and Dense LIDAR Data. IROS 2014.
59	ACF-MR			46.23 %	58.82 %	42.10 %	0.6 s	1 core @ 3.5 Ghz (C/C++)
R. Rajaram, E. Ohn-Bar and M. Trivedi: Looking at Pedestrians at Different Scales: A Multi-resolution Approach and Evaluations. T-ITS 2016.
60	HA-SSVM			45.51 %	56.36 %	41.08 %	21 s	1 core @ >3.5 Ghz (Matlab + C/C++)
J. Xu, S. Ramos, D. Vázquez and A. López: Hierarchical Adaptive Structural SVM for Domain Adaptation. IJCV 2016.
61	DPM-VOC+VP			44.86 %	59.48 %	40.37 %	8 s	1 core @ 2.5 Ghz (C/C++)
B. Pepik, M. Stark, P. Gehler and B. Schiele: Multi-view and 3D Deformable Part Models. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) 2015.
62	ACF-SC			44.49 %	51.53 %	40.38 %	<0.3 s	1 core @ >3.5 Ghz (Matlab + C/C++)
C. Cadena, A. Dick and I. Reid: A Fast, Modular Scene Understanding System using Context-Aware Object Detection. Robotics and Automation (ICRA), 2015 IEEE International Conference on 2015.
63	SquaresICF		code	44.42 %	57.33 %	40.08 %	1 s	GPU @ >3.5 Ghz (C/C++)
R. Benenson, M. Mathias, T. Tuytelaars and L. Gool: Seeking the strongest rigid detector. CVPR 2013.
64	AR-FCN			43.88 %	53.16 %	35.58 %	0.19 s	GPU @ 2.5 Ghz (C/C++)
65	QHY			43.42 %	60.08 %	42.31 %	0.1 s	1 core @ 2.5 Ghz (C/C++)
66	SubCat			42.34 %	54.67 %	37.95 %	1.2 s	6 cores @ 2.5 Ghz (Matlab + C/C++)
E. Ohn-Bar and M. Trivedi: Fast and Robust Object Detection Using Visual Subcategories. Computer Vision and Pattern Recognition Workshops Mobile Vision 2014.
67	HL			42.31 %	58.55 %	34.87 %	0.16 s	1 core @ 2.5 Ghz (C/C++)
68	RCNN			42.16 %	58.37 %	34.88 %	0.08 s	GPU @ 2.5 Ghz (Python + C/C++)
69	Fast-RCNN-SS			41.57 %	52.68 %	35.25 %	1 s	GPU @ 2.0 Ghz (Matlab + C/C++)
70	NMRDO			40.59 %	54.87 %	39.75 %	0.1 s	GPU @ 2.5 Ghz (Python + C/C++)
71	ACFK		code	40.23 %	48.83 %	33.57 %	0.07 s	1 core @ >3.5 Ghz (C/C++)
72	ACF			39.81 %	44.49 %	37.21 %	0.2 s	1 core @ >3.5 Ghz (Matlab + C/C++)
P. Doll\'ar, R. Appel, S. Belongie and P. Perona: Fast Feature Pyramids for Object Detection. PAMI 2014. P. Doll\'ar: Piotr's Image and Video Matlab Toolbox (PMT). .
73	ACF_M			39.36 %	47.74 %	35.95 %	0.1 s	1 core @ 2.5 Ghz (C/C++)
74	LSVM-MDPM-sv			39.36 %	51.75 %	35.95 %	10 s	4 cores @ 3.0 Ghz (C/C++)
P. Felzenszwalb, R. Girshick, D. McAllester and D. Ramanan: Object Detection with Discriminatively Trained Part-Based Models. PAMI 2010. A. Geiger, C. Wojek and R. Urtasun: Joint 3D Estimation of Objects and Scene Layout. NIPS 2011.
75	PCNN			39.07 %	53.37 %	37.91 %	1 s	1 core @ 2.5 Ghz (C/C++)
76	CNN			38.98 %	52.84 %	38.31 %	1 s	1 core @ 2.5 Ghz (C/C++)
77	LSVM-MDPM-us		code	38.35 %	45.50 %	34.78 %	10 s	4 cores @ 3.0 Ghz (C/C++)
P. Felzenszwalb, R. Girshick, D. McAllester and D. Ramanan: Object Detection with Discriminatively Trained Part-Based Models. PAMI 2010.
78	Vote3D			35.74 %	44.48 %	33.72 %	0.5 s	4 cores @ 2.8 Ghz (C/C++)
D. Wang and I. Posner: Voting for Voting in Online Point Cloud Object Detection. Proceedings of Robotics: Science and Systems 2015.
79	mBoW			31.37 %	44.28 %	30.62 %	10 s	1 core @ 2.5 Ghz (C/C++)
J. Behley, V. Steinhage and A. Cremers: Laser-based Segment Classification Using a Mixture of Bag-of-Words. Proc. of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2013.
80	DPM-C8B1			29.03 %	38.96 %	25.61 %	15 s	4 cores @ 2.5 Ghz (Matlab + C/C++)
J. Yebes, L. Bergasa and M. García-Garrido: Visual Object Recognition with 3D-Aware Features in KITTI Urban Scenes. Sensors 2015. J. Yebes, L. Bergasa, R. Arroyo and A. Lázaro: Supervised learning and evaluation of KITTI's cars detector with DPM. IV 2014.
81	YOLO			24.35 %	25.63 %	17.50 %	0.03 s	GPU @ 1.0 Ghz (C/C++)
82	R-CNN_VGG			23.14 %	29.06 %	22.15 %	10 s	GPU @ 2.5 Ghz (Matlab + C/C++)
83	YOLOv2		code	16.19 %	20.64 %	15.43 %	0.02 s	GPU @ 3.5 Ghz (C/C++)
J. Redmon, S. Divvala, R. Girshick and A. Farhadi: You only look once: Unified, real-time object detection. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2016. J. Redmon and A. Farhadi: YOLO9000: Better, Faster, Stronger. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2017.
84	BIP-HETERO			13.38 %	14.85 %	13.25 %	~2 s	1 core @ 2.5 Ghz (C/C++)
A. Mekonnen, F. Lerasle, A. Herbulot and C. Briand: People Detection with Heterogeneous Features and Explicit Optimization on Computation Time. Pattern Recognition (ICPR), 2014 22nd International Conference on 2014.

Cyclist

	Method	Setting	Code	Moderate	Easy	Hard	Runtime	Environment
1	RRC		code	76.47 %	84.71 %	65.46 %	3.6 s	GPU @ 2.5 Ghz (Python + C/C++)
J. Ren, X. Chen, J. Liu, W. Sun, J. Pang, Q. Yan, Y. Tai and L. Xu: Accurate Single Stage Detector Using Recurrent Rolling Convolution. CVPR 2017.
2	Pie			76.25 %	84.62 %	67.57 %	1.2 s	1 core @ 2.5 Ghz (C/C++)
3	SAIT			76.13 %	83.88 %	66.60 %	0.15 s	GPU @ >3.5 Ghz (Python + C/C++)
4	TiCNN			75.83 %	84.28 %	66.50 %	0.5 s	GPU @ 2.5 Ghz (Matlab + C/C++)
5	TuSimple			75.59 %	84.15 %	66.35 %	1.6 s	GPU @ 2.5 Ghz (Python + C/C++)
6	Allspark			75.47 %	84.88 %	66.35 %	0.7 s	GPU @ 2.5 Ghz (C/C++)
7	MS-CNN		code	75.46 %	84.06 %	66.07 %	0.4 s	GPU @ 2.5 Ghz (C/C++)
Z. Cai, Q. Fan, R. Feris and N. Vasconcelos: A Unified Multi-scale Deep Convolutional Neural Network for Fast Object Detection. ECCV 2016.
8	Deep3DBox			74.16 %	83.94 %	64.84 %	1.5 s	GPU @ 2.5 Ghz (C/C++)
A. Mousavian, D. Anguelov, J. Flynn and J. Kosecka: 3D Bounding Box Estimation Using Deep Learning and Geometry. CVPR 2017.
9	SDP+RPN			73.74 %	81.37 %	65.31 %	0.4 s	GPU @ 2.5 Ghz (Python + C/C++)
F. Yang, W. Choi and Y. Lin: Exploit All the Layers: Fast and Accurate CNN Object Detector with Scale Dependent Pooling and Cascaded Rejection Classifiers. Proceedings of the IEEE International Conference on Computer Vision and Pattern Recognition 2016. S. Ren, K. He, R. Girshick and J. Sun: Faster R-CNN: Towards real-time object detection with region proposal networks. Advances in Neural Information Processing Systems 2015.
10	sensekitti			72.56 %	82.39 %	64.00 %	4.5 s	GPU @ 2.5 Ghz (Python + C/C++)
11	SubCNN			71.06 %	79.48 %	62.68 %	2 s	GPU @ 3.5 Ghz (Python + C/C++)
Y. Xiang, W. Choi, Y. Lin and S. Savarese: Subcategory-aware Convolutional Neural Networks for Object Proposals and Detection. IEEE Winter Conference on Applications of Computer Vision (WACV) 2017.
12	uickitti			70.90 %	78.40 %	62.54 %	1.5 s	GPU @ 2.5 Ghz (C/C++)
13	DJML			70.26 %	79.41 %	61.84 %	2.4 s	GPU @ 2.5 Ghz (Python + C/C++)
14	3DOP		code	68.94 %	78.39 %	61.37 %	3s	GPU @ 2.5 Ghz (Matlab + C/C++)
X. Chen, K. Kundu, Y. Zhu, A. Berneshawi, H. Ma, S. Fidler and R. Urtasun: 3D Object Proposals for Accurate Object Class Detection. NIPS 2015.
15	Pose-RCNN			68.79 %	80.79 %	60.40 %	2 s	>8 cores @ 2.5 Ghz (Python)
16	Re-3DOP			68.31 %	78.08 %	60.73 %	3 s	1 core @ 2.5 Ghz (C/C++)
17	Vote3Deep			67.88 %	79.92 %	62.98 %	1.5 s	4 cores @ 2.5 Ghz (C/C++)
M. Engelcke, D. Rao, D. Zeng Wang, C. Hay Tong and I. Posner: Vote3Deep: Fast Object Detection in 3D Point Clouds Using Efficient Convolutional Neural Networks. ArXiv e-prints 2016.
18	IVA		code	67.47 %	80.17 %	59.66 %	0.4 s	GPU @ 2.5 Ghz (C/C++)
Y. Zhu, J. Wang, C. Zhao, H. Guo and H. Lu: Scale-adaptive Deconvolutional Regression Network for Pedestrian Detection. ACCV 2016. S. Ren, K. He, R. Girshick and J. Sun: Faster R-CNN: Towards real-time object detection with region proposal networks. Advances in neural information processing systems 2015.
19	Mono3D		code	66.36 %	76.04 %	58.87 %	4.2 s	GPU @ 2.5 Ghz (Matlab + C/C++)
X. Chen, K. Kundu, Z. Zhang, H. Ma, S. Fidler and R. Urtasun: Monocular 3D Object Detection for Autonomous Driving. CVPR 2016.
20	DeepStereoOP			65.84 %	79.58 %	57.90 %	3.4 s	GPU @ 3.5 Ghz (Matlab + C/C++)
C. Pham and J. Jeon: Robust Object Proposals Re-ranking for Object Detection in Autonomous Driving Using Convolutional Neural Networks. Signal Processing: Image Communiation 2017.
21	HM_SSD_RCNN			65.57 %	78.24 %	57.66 %	0.15 s	1 core @ 2.5 Ghz (C/C++)
22	HSR2			64.92 %	74.75 %	57.49 %	0.15 s	1 core @ 2.5 Ghz (C/C++)
23	tbd			63.81 %	75.49 %	56.26 %	1 s	1 core @ 2.5 Ghz (C/C++)
24	Faster R-CNN		code	63.35 %	72.26 %	55.90 %	2 s	GPU @ 3.5 Ghz (Python + C/C++)
S. Ren, K. He, R. Girshick and J. Sun: Faster R-CNN: Towards Real- Time Object Detection with Region Proposal Networks. NIPS 2015.
25	WRInception			62.85 %	77.21 %	55.80 %	0.06 s	GPU @ 2.5 Ghz (C/C++)
26	SDP+CRC (ft)			61.31 %	74.08 %	53.97 %	0.6 s	GPU @ 2.5 Ghz (C/C++)
F. Yang, W. Choi and Y. Lin: Exploit All the Layers: Fast and Accurate CNN Object Detector with Scale Dependent Pooling and Cascaded Rejection Classifiers. Proceedings of the IEEE International Conference on Computer Vision and Pattern Recognition 2016.
27	IVA		code	61.11 %	69.26 %	54.28 %	1 s	GPU @ 2.5 Ghz (Matlab + C/C++)
28	Regionlets			58.72 %	70.41 %	51.83 %	1 s	>8 cores @ 2.5 Ghz (C/C++)
X. Wang, M. Yang, S. Zhu and Y. Lin: Regionlets for Generic Object Detection. T-PAMI 2015. W. Zou, X. Wang, M. Sun and Y. Lin: Generic Object Detection with Dense Neural Patterns and Regionlets. British Machine Vision Conference 2014. C. Long, X. Wang, G. Hua, M. Yang and Y. Lin: Accurate Object Detection with Location Relaxation and Regionlets Relocalization. Asian Conference on Computer Vision 2014.
29	PNET			58.60 %	74.27 %	51.66 %	0.1 s	GPU @ 2.5 Ghz (Python + C/C++)
30	ANM			53.04 %	71.56 %	46.38 %	0.05 s	GPU @ 2.5 Ghz (C/C++)
31	ZGC			48.02 %	64.32 %	40.71 %	0.12 s	1 core @ 2.5 Ghz (C/C++)
32	FD2			44.13 %	61.84 %	40.35 %	0.01 s	GPU @ >3.5 Ghz (Python + C/C++)
33	maxFtr+ROI			43.58 %	48.38 %	38.73 %	0.25 s	4 cores @ 2.5 Ghz (C/C++)
W. Tian and M. Lauer: Detection and Orientation Estimation for Cyclists by Max Pooled Features. International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP) 2017.
34	MV-RGBD-RF			42.61 %	52.97 %	37.42 %	4 s	4 cores @ 2.5 Ghz (C/C++)
A. Gonzalez, D. Vazquez, A. Lopez and J. Amores: On-Board Object Detection: Multicue, Multimodal, and Multiview Random Forest of Local Experts.. IEEE Trans. on Cybernetics 2016. A. Gonzalez, G. Villalonga, J. Xu, D. Vazquez, J. Amores and A. Lopez: Multiview Random Forest of Local Experts Combining RGB and LIDAR data for Pedestrian Detection. IEEE Intelligent Vehicles Symposium (IV) 2015.
35	QHY			42.30 %	59.15 %	41.28 %	0.1 s	1 core @ 2.5 Ghz (C/C++)
36	NMF-CNN			42.11 %	56.14 %	37.45 %	0.1 s	GPU @ 2.5 Ghz (Matlab + C/C++)
37	AR-FCN			41.83 %	58.08 %	33.99 %	0.19 s	GPU @ 2.5 Ghz (C/C++)
38	RCNN			40.36 %	56.67 %	33.07 %	0.08 s	GPU @ 2.5 Ghz (Python + C/C++)
39	HL			39.07 %	54.45 %	32.63 %	0.16 s	1 core @ 2.5 Ghz (C/C++)
40	pAUCEnsT			38.03 %	51.62 %	33.38 %	60 s	1 core @ 2.5 Ghz (Matlab + C/C++)
S. Paisitkriangkrai, C. Shen and A. Hengel: Pedestrian Detection with Spatially Pooled Features and Structured Ensemble Learning. arXiv 2014.
41	FD			36.58 %	50.51 %	32.31 %	0.01 s	GPU @ >3.5 Ghz (Python)
42	NMRDO			33.43 %	45.02 %	27.79 %	0.1 s	GPU @ 2.5 Ghz (Python + C/C++)
43	Vote3D			31.24 %	41.43 %	28.60 %	0.5 s	4 cores @ 2.8 Ghz (C/C++)
D. Wang and I. Posner: Voting for Voting in Online Point Cloud Object Detection. Proceedings of Robotics: Science and Systems 2015.
44	DPM-VOC+VP			31.08 %	42.43 %	28.23 %	8 s	1 core @ 2.5 Ghz (C/C++)
B. Pepik, M. Stark, P. Gehler and B. Schiele: Multi-view and 3D Deformable Part Models. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) 2015.
45	LSVM-MDPM-us		code	29.88 %	38.84 %	27.31 %	10 s	4 cores @ 3.0 Ghz (C/C++)
P. Felzenszwalb, R. Girshick, D. McAllester and D. Ramanan: Object Detection with Discriminatively Trained Part-Based Models. PAMI 2010.
46	LSVM-MDPM-sv			29.24 %	37.71 %	27.52 %	10 s	4 cores @ 3.0 Ghz (C/C++)
P. Felzenszwalb, R. Girshick, D. McAllester and D. Ramanan: Object Detection with Discriminatively Trained Part-Based Models. PAMI 2010. A. Geiger, C. Wojek and R. Urtasun: Joint 3D Estimation of Objects and Scene Layout. NIPS 2011.
47	DPM-C8B1			29.04 %	43.49 %	26.20 %	15 s	4 cores @ 2.5 Ghz (Matlab + C/C++)
J. Yebes, L. Bergasa and M. García-Garrido: Visual Object Recognition with 3D-Aware Features in KITTI Urban Scenes. Sensors 2015. J. Yebes, L. Bergasa, R. Arroyo and A. Lázaro: Supervised learning and evaluation of KITTI's cars detector with DPM. IV 2014.
48	R-CNN_VGG			28.76 %	37.74 %	25.84 %	10 s	GPU @ 2.5 Ghz (Matlab + C/C++)
49	ACF_M			27.50 %	35.04 %	26.21 %	0.1 s	1 core @ 2.5 Ghz (C/C++)
50	mBoW			21.62 %	28.00 %	20.93 %	10 s	1 core @ 2.5 Ghz (C/C++)
J. Behley, V. Steinhage and A. Cremers: Laser-based Segment Classification Using a Mixture of Bag-of-Words. Proc. of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2013.
51	YOLO			13.96 %	17.93 %	13.83 %	0.03 s	GPU @ 1.0 Ghz (C/C++)
52	YOLOv2		code	4.55 %	4.55 %	4.55 %	0.02 s	GPU @ 3.5 Ghz (C/C++)
J. Redmon, S. Divvala, R. Girshick and A. Farhadi: You only look once: Unified, real-time object detection. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2016. J. Redmon and A. Farhadi: YOLO9000: Better, Faster, Stronger. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2017.

Cars

	Method	Setting	Code	Moderate	Easy	Hard	Runtime	Environment
1	Deep MANTA			89.86 %	95.72 %	80.39 %	0.7 s	GPU @ 2.5 Ghz (Python + C/C++)
F. Chabot, M. Chaouch, J. Rabarisoa, C. Teulière and T. Chateau: Deep MANTA: A Coarse-to-fine Many-Task Network for joint 2D and 3D vehicle analysis from monocular image. CVPR 2017.
2	uickitti			88.79 %	90.70 %	78.84 %	1.5 s	GPU @ 2.5 Ghz (C/C++)
3	Deep3DBox			88.75 %	92.90 %	76.76 %	1.5 s	GPU @ 2.5 Ghz (C/C++)
A. Mousavian, D. Anguelov, J. Flynn and J. Kosecka: 3D Bounding Box Estimation Using Deep Learning and Geometry. CVPR 2017.
4	SubCNN			88.62 %	90.67 %	78.68 %	2 s	GPU @ 3.5 Ghz (Python + C/C++)
Y. Xiang, W. Choi, Y. Lin and S. Savarese: Subcategory-aware Convolutional Neural Networks for Object Proposals and Detection. IEEE Winter Conference on Applications of Computer Vision (WACV) 2017.
5	DJML			88.27 %	90.57 %	78.30 %	2.4 s	GPU @ 2.5 Ghz (Python + C/C++)
6	DeepStereoOP			86.86 %	92.04 %	77.34 %	3.4 s	GPU @ 3.5 Ghz (Matlab + C/C++)
C. Pham and J. Jeon: Robust Object Proposals Re-ranking for Object Detection in Autonomous Driving Using Convolutional Neural Networks. Signal Processing: Image Communiation 2017.
7	Mono3D		code	86.62 %	91.01 %	76.84 %	4.2 s	GPU @ 2.5 Ghz (Matlab + C/C++)
X. Chen, K. Kundu, Z. Zhang, H. Ma, S. Fidler and R. Urtasun: Monocular 3D Object Detection for Autonomous Driving. CVPR 2016.
8	3DOP		code	86.10 %	91.44 %	76.52 %	3s	GPU @ 2.5 Ghz (Matlab + C/C++)
X. Chen, K. Kundu, Y. Zhu, A. Berneshawi, H. Ma, S. Fidler and R. Urtasun: 3D Object Proposals for Accurate Object Class Detection. NIPS 2015.
9	Pose-RCNN			75.41 %	88.34 %	66.07 %	2 s	>8 cores @ 2.5 Ghz (Python)
10	XXX			75.22 %	84.09 %	67.87 %	>5 s	1 core @ 2.5 Ghz (C/C++)
11	3DVP		code	74.59 %	86.92 %	64.11 %	40 s	8 cores @ 3.5 Ghz (Matlab + C/C++)
Y. Xiang, W. Choi, Y. Lin and S. Savarese: Data-Driven 3D Voxel Patterns for Object Category Recognition. IEEE Conference on Computer Vision and Pattern Recognition 2015.
12	SubCat		code	74.42 %	83.41 %	58.83 %	0.7 s	6 cores @ 3.5 Ghz (Matlab + C/C++)
E. Ohn-Bar and M. Trivedi: Learning to Detect Vehicles by Clustering Appearance Patterns. T-ITS 2015.
13	BdCost48LDCF			65.63 %	76.54 %	54.56 %	5 s	1 core @ 2.5 Ghz (C/C++)
14	BdCost48-25C			65.61 %	76.95 %	55.02 %	4 s	1 core @ 2.5 Ghz (C/C++)
15	OC-DPM			64.42 %	73.50 %	52.40 %	10 s	8 cores @ 2.5 Ghz (Matlab)
B. Pepik, M. Stark, P. Gehler and B. Schiele: Occlusion Patterns for Object Class Detection. IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2013.
16	AOG-View			63.31 %	76.70 %	50.34 %	3 s	1 core @ 2.5 Ghz (Matlab, C/C++)
B. Li, T. Wu and S. Zhu: Integrating Context and Occlusion for Car Detection by Hierarchical And-Or Model. ECCV 2014.
17	DPM-VOC+VP			61.84 %	72.28 %	46.54 %	8 s	1 core @ 2.5 Ghz (C/C++)
B. Pepik, M. Stark, P. Gehler and B. Schiele: Multi-view and 3D Deformable Part Models. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) 2015.
18	NMRDO			59.55 %	70.51 %	51.91 %	0.1 s	GPU @ 2.5 Ghz (Python + C/C++)
19	LSVM-MDPM-sv			56.69 %	70.86 %	45.91 %	10 s	4 cores @ 3.0 Ghz (C/C++)
P. Felzenszwalb, R. Girshick, D. McAllester and D. Ramanan: Object Detection with Discriminatively Trained Part-Based Models. PAMI 2010. A. Geiger, C. Wojek and R. Urtasun: Joint 3D Estimation of Objects and Scene Layout. NIPS 2011.
20	GVPL			54.32 %	63.14 %	46.24 %	1 s	8 cores @ 2.5 Ghz (Matlab + C/C++)
21	VeloFCN			52.84 %	70.58 %	46.14 %	1 s	GPU @ 2.5 Ghz (Python + C/C++)
B. Li, T. Zhang and T. Xia: Vehicle Detection from 3D Lidar Using Fully Convolutional Network. RSS 2016 .
22	DPM-C8B1			50.32 %	59.51 %	39.22 %	15 s	4 cores @ 2.5 Ghz (Matlab + C/C++)
J. Yebes, L. Bergasa and M. García-Garrido: Visual Object Recognition with 3D-Aware Features in KITTI Urban Scenes. Sensors 2015. J. Yebes, L. Bergasa, R. Arroyo and A. Lázaro: Supervised learning and evaluation of KITTI's cars detector with DPM. IV 2014.
23	HSR2			45.74 %	49.18 %	40.91 %	0.15 s	1 core @ 2.5 Ghz (C/C++)
24	Allspark			45.33 %	47.24 %	39.81 %	0.7 s	GPU @ 2.5 Ghz (C/C++)
25	WRInception			45.05 %	46.75 %	40.69 %	0.06 s	GPU @ 2.5 Ghz (C/C++)
26	HM_SSD_RCNN			44.61 %	48.34 %	40.47 %	0.15 s	1 core @ 2.5 Ghz (C/C++)
27	sensekitti			44.57 %	47.32 %	41.44 %	4.5 s	GPU @ 2.5 Ghz (Python + C/C++)
28	DuEye			42.06 %	39.97 %	39.28 %	4 s	GPU @ 2.5 Ghz (C/C++)
29	FD			39.81 %	44.14 %	34.80 %	0.01 s	GPU @ >3.5 Ghz (Python)
30	FD2			39.50 %	47.54 %	35.39 %	0.01 s	GPU @ >3.5 Ghz (Python + C/C++)
31	CPCD			39.02 %	37.90 %	34.23 %	3 s	1 core @ 2.5 Ghz (C/C++)
32	Re-3DOP			38.46 %	37.92 %	33.91 %	3 s	1 core @ 2.5 Ghz (C/C++)
33	UI			38.10 %	38.55 %	34.44 %	0.4 s	GPU @ 2.5 Ghz (C/C++)
34	Direwolf			36.93 %	39.69 %	33.38 %	0.5 s	GPU @ 2.5 Ghz (C/C++)
35	ZGC			36.55 %	45.25 %	32.07 %	0.12 s	1 core @ 2.5 Ghz (C/C++)
36	SYVO			36.28 %	36.76 %	29.56 %	0.13 s	GPU @ 2.5 Ghz (C/C++)
37	QHY			36.24 %	45.66 %	31.53 %	0.1 s	1 core @ 2.5 Ghz (C/C++)
38	HL			34.92 %	43.13 %	27.91 %	0.16 s	1 core @ 2.5 Ghz (C/C++)
39	ANM			34.79 %	35.30 %	31.75 %	0.05 s	GPU @ 2.5 Ghz (C/C++)
40	SceneNet			31.98 %	36.26 %	28.41 %	0.03 s	GPU @ 2.5 Ghz (C/C++)
41	AOG		code	30.77 %	33.79 %	24.75 %	3 s	4 cores @ 2.5 Ghz (Matlab)
T. Wu, B. Li and S. Zhu: Learning And-Or Models to Represent Context and Occlusion for Car Detection and Viewpoint Estimation. TPAMI 2016. B. Li, T. Wu and S. Zhu: Integrating Context and Occlusion for Car Detection by Hierarchical And-Or Model. ECCV 2014.
42	FCNN			28.85 %	35.19 %	25.25 %	0.1 s	1 core @ 2.5 Ghz (C/C++)
43	NMF-CNN			26.37 %	32.69 %	20.48 %	0.1 s	GPU @ 2.5 Ghz (Matlab + C/C++)
44	CSoR		code	25.38 %	33.97 %	21.95 %	3.5 s	4 cores @ >3.5 Ghz (Python + C/C++)
L. Plotkin: PyDriver: Entwicklung eines Frameworks für räumliche Detektion und Klassifikation von Objekten in Fahrzeugumgebung. 2015.
45	SubCat48LDCF			24.37 %	26.21 %	18.74 %	5 s	1 core @ 2.5 Ghz (Matlab + C/C++)
46	frd			22.23 %	27.63 %	19.39 %	2 s	1 core @ 2.5 Ghz (C/C++)

Pedestrians

	Method	Setting	Code	Moderate	Easy	Hard	Runtime	Environment
1	uickitti			66.83 %	78.93 %	62.06 %	1.5 s	GPU @ 2.5 Ghz (C/C++)
2	SubCNN			66.28 %	78.45 %	61.36 %	2 s	GPU @ 3.5 Ghz (Python + C/C++)
Y. Xiang, W. Choi, Y. Lin and S. Savarese: Subcategory-aware Convolutional Neural Networks for Object Proposals and Detection. IEEE Winter Conference on Applications of Computer Vision (WACV) 2017.
3	Pose-RCNN			59.90 %	73.95 %	54.27 %	2 s	>8 cores @ 2.5 Ghz (Python)
4	3DOP		code	59.80 %	72.94 %	57.03 %	3s	GPU @ 2.5 Ghz (Matlab + C/C++)
X. Chen, K. Kundu, Y. Zhu, A. Berneshawi, H. Ma, S. Fidler and R. Urtasun: 3D Object Proposals for Accurate Object Class Detection. NIPS 2015.
5	DeepStereoOP			59.28 %	72.82 %	56.85 %	3.4 s	GPU @ 3.5 Ghz (Matlab + C/C++)
C. Pham and J. Jeon: Robust Object Proposals Re-ranking for Object Detection in Autonomous Driving Using Convolutional Neural Networks. Signal Processing: Image Communiation 2017.
6	Mono3D		code	58.15 %	71.15 %	54.94 %	4.2 s	GPU @ 2.5 Ghz (Matlab + C/C++)
X. Chen, K. Kundu, Z. Zhang, H. Ma, S. Fidler and R. Urtasun: Monocular 3D Object Detection for Autonomous Driving. CVPR 2016.
7	DJML			58.12 %	69.35 %	52.62 %	2.4 s	GPU @ 2.5 Ghz (Python + C/C++)
8	DPM-VOC+VP			39.83 %	53.55 %	35.73 %	8 s	1 core @ 2.5 Ghz (C/C++)
B. Pepik, M. Stark, P. Gehler and B. Schiele: Multi-view and 3D Deformable Part Models. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) 2015.
9	Allspark			38.99 %	43.71 %	36.22 %	0.7 s	GPU @ 2.5 Ghz (C/C++)
10	sensekitti			37.50 %	43.27 %	35.10 %	4.5 s	GPU @ 2.5 Ghz (Python + C/C++)
11	Re-3DOP			36.28 %	44.46 %	34.34 %	3 s	1 core @ 2.5 Ghz (C/C++)
12	ACF_M			35.49 %	43.58 %	32.42 %	0.1 s	1 core @ 2.5 Ghz (C/C++)
13	LSVM-MDPM-sv			35.49 %	47.00 %	32.42 %	10 s	4 cores @ 3.0 Ghz (C/C++)
P. Felzenszwalb, R. Girshick, D. McAllester and D. Ramanan: Object Detection with Discriminatively Trained Part-Based Models. PAMI 2010. A. Geiger, C. Wojek and R. Urtasun: Joint 3D Estimation of Objects and Scene Layout. NIPS 2011.
14	WRInception			35.12 %	40.32 %	32.48 %	0.06 s	GPU @ 2.5 Ghz (C/C++)
15	HM_SSD_RCNN			34.37 %	41.89 %	30.73 %	0.15 s	1 core @ 2.5 Ghz (C/C++)
16	SubCat			34.18 %	44.32 %	30.76 %	1.2 s	6 cores @ 2.5 Ghz (Matlab + C/C++)
E. Ohn-Bar and M. Trivedi: Fast and Robust Object Detection Using Visual Subcategories. Computer Vision and Pattern Recognition Workshops Mobile Vision 2014.
17	HSR2			33.86 %	41.48 %	32.48 %	0.15 s	1 core @ 2.5 Ghz (C/C++)
18	Tx			33.82 %	39.55 %	30.97 %	2 s	GPU @ 2.5 Ghz (Matlab + C/C++)
19	RB			33.70 %	43.32 %	30.29 %	0.6 s	GPU @ 2.5 Ghz (Matlab + C/C++)
20	NMRDO			33.06 %	44.55 %	31.83 %	0.1 s	GPU @ 2.5 Ghz (Python + C/C++)
21	SSD1			32.66 %	41.64 %	30.50 %	0.255 s	GPU @ 2.5 Ghz (python+ C/C++)
22	RPN+BF		code	32.55 %	40.91 %	29.52 %	0.6 s	GPU @ 2.5 Ghz (Matlab + C/C++)
L. Zhang, L. Lin, X. Liang and K. He: Is Faster R-CNN Doing Well for Pedestrian Detection?. ECCV 2016.
23	NMF-CNN			30.94 %	40.13 %	28.65 %	0.1 s	GPU @ 2.5 Ghz (Matlab + C/C++)
24	FD2			28.57 %	35.54 %	26.00 %	0.01 s	GPU @ >3.5 Ghz (Python + C/C++)
25	ANM			28.47 %	38.07 %	27.69 %	0.05 s	GPU @ 2.5 Ghz (C/C++)
26	ACF			28.46 %	35.69 %	26.18 %	1 s	1 core @ 3.5 Ghz (Matlab + C/C++)
P. Doll\'ar, R. Appel, S. Belongie and P. Perona: Fast Feature Pyramids for Object Detection. PAMI 2014.
27	FD			27.77 %	33.18 %	25.06 %	0.01 s	GPU @ >3.5 Ghz (Python)
28	ZGC			26.42 %	34.53 %	22.57 %	0.12 s	1 core @ 2.5 Ghz (C/C++)
29	HL			24.21 %	32.28 %	20.43 %	0.16 s	1 core @ 2.5 Ghz (C/C++)
30	DPM-C8B1			23.37 %	31.08 %	20.72 %	15 s	4 cores @ 2.5 Ghz (Matlab + C/C++)
J. Yebes, L. Bergasa and M. García-Garrido: Visual Object Recognition with 3D-Aware Features in KITTI Urban Scenes. Sensors 2015. J. Yebes, L. Bergasa, R. Arroyo and A. Lázaro: Supervised learning and evaluation of KITTI's cars detector with DPM. IV 2014.
31	ACF-MR			23.18 %	29.33 %	21.00 %	0.6 s	1 core @ 3.5 Ghz (C/C++)
R. Rajaram, E. Ohn-Bar and M. Trivedi: Looking at Pedestrians at Different Scales: A Multi-resolution Approach and Evaluations. T-ITS 2016.
32	QHY			21.79 %	30.54 %	21.41 %	0.1 s	1 core @ 2.5 Ghz (C/C++)

Cyclists

	Method	Setting	Code	Moderate	Easy	Hard	Runtime	Environment
1	uickitti			63.69 %	71.04 %	56.34 %	1.5 s	GPU @ 2.5 Ghz (C/C++)
2	SubCNN			63.65 %	72.00 %	56.32 %	2 s	GPU @ 3.5 Ghz (Python + C/C++)
Y. Xiang, W. Choi, Y. Lin and S. Savarese: Subcategory-aware Convolutional Neural Networks for Object Proposals and Detection. IEEE Winter Conference on Applications of Computer Vision (WACV) 2017.
3	Pose-RCNN			62.87 %	75.49 %	55.47 %	2 s	>8 cores @ 2.5 Ghz (Python)
4	Deep3DBox			59.87 %	69.16 %	52.50 %	1.5 s	GPU @ 2.5 Ghz (C/C++)
A. Mousavian, D. Anguelov, J. Flynn and J. Kosecka: 3D Bounding Box Estimation Using Deep Learning and Geometry. CVPR 2017.
5	DJML			59.71 %	69.59 %	52.89 %	2.4 s	GPU @ 2.5 Ghz (Python + C/C++)
6	3DOP		code	58.68 %	70.13 %	52.35 %	3s	GPU @ 2.5 Ghz (Matlab + C/C++)
X. Chen, K. Kundu, Y. Zhu, A. Berneshawi, H. Ma, S. Fidler and R. Urtasun: 3D Object Proposals for Accurate Object Class Detection. NIPS 2015.
7	DeepStereoOP			55.69 %	69.20 %	48.95 %	3.4 s	GPU @ 3.5 Ghz (Matlab + C/C++)
C. Pham and J. Jeon: Robust Object Proposals Re-ranking for Object Detection in Autonomous Driving Using Convolutional Neural Networks. Signal Processing: Image Communiation 2017.
8	Mono3D		code	54.97 %	65.56 %	48.77 %	4.2 s	GPU @ 2.5 Ghz (Matlab + C/C++)
X. Chen, K. Kundu, Z. Zhang, H. Ma, S. Fidler and R. Urtasun: Monocular 3D Object Detection for Autonomous Driving. CVPR 2016.
9	Allspark			43.03 %	49.80 %	37.00 %	0.7 s	GPU @ 2.5 Ghz (C/C++)
10	sensekitti			42.12 %	46.41 %	36.62 %	4.5 s	GPU @ 2.5 Ghz (Python + C/C++)
11	maxFtr+ROI			38.28 %	41.82 %	34.27 %	0.25 s	4 cores @ 2.5 Ghz (C/C++)
W. Tian and M. Lauer: Detection and Orientation Estimation for Cyclists by Max Pooled Features. International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP) 2017.
12	HSR2			36.83 %	41.73 %	32.26 %	0.15 s	1 core @ 2.5 Ghz (C/C++)
13	HM_SSD_RCNN			36.26 %	44.69 %	31.31 %	0.15 s	1 core @ 2.5 Ghz (C/C++)
14	WRInception			34.20 %	41.42 %	29.61 %	0.06 s	GPU @ 2.5 Ghz (C/C++)
15	Re-3DOP			29.60 %	31.39 %	27.37 %	3 s	1 core @ 2.5 Ghz (C/C++)
16	ZGC			26.50 %	36.20 %	22.24 %	0.12 s	1 core @ 2.5 Ghz (C/C++)
17	FD2			24.55 %	35.33 %	21.81 %	0.01 s	GPU @ >3.5 Ghz (Python + C/C++)
18	QHY			23.90 %	33.59 %	22.74 %	0.1 s	1 core @ 2.5 Ghz (C/C++)
19	NMRDO			23.53 %	31.47 %	19.81 %	0.1 s	GPU @ 2.5 Ghz (Python + C/C++)
20	DPM-VOC+VP			23.17 %	30.52 %	21.58 %	8 s	1 core @ 2.5 Ghz (C/C++)
B. Pepik, M. Stark, P. Gehler and B. Schiele: Multi-view and 3D Deformable Part Models. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) 2015.
21	LSVM-MDPM-sv			23.14 %	28.89 %	22.28 %	10 s	4 cores @ 3.0 Ghz (C/C++)
P. Felzenszwalb, R. Girshick, D. McAllester and D. Ramanan: Object Detection with Discriminatively Trained Part-Based Models. PAMI 2010. A. Geiger, C. Wojek and R. Urtasun: Joint 3D Estimation of Objects and Scene Layout. NIPS 2011.
22	ANM			22.82 %	30.83 %	20.15 %	0.05 s	GPU @ 2.5 Ghz (C/C++)
23	ACF_M			22.07 %	27.54 %	21.45 %	0.1 s	1 core @ 2.5 Ghz (C/C++)
24	HL			21.40 %	29.81 %	17.63 %	0.16 s	1 core @ 2.5 Ghz (C/C++)
25	FD			21.35 %	30.37 %	18.20 %	0.01 s	GPU @ >3.5 Ghz (Python)
26	DPM-C8B1			19.25 %	27.25 %	17.95 %	15 s	4 cores @ 2.5 Ghz (Matlab + C/C++)
J. Yebes, L. Bergasa and M. García-Garrido: Visual Object Recognition with 3D-Aware Features in KITTI Urban Scenes. Sensors 2015. J. Yebes, L. Bergasa, R. Arroyo and A. Lázaro: Supervised learning and evaluation of KITTI's cars detector with DPM. IV 2014.
27	NMF-CNN			16.77 %	21.97 %	15.09 %	0.1 s	GPU @ 2.5 Ghz (Matlab + C/C++)

• PASCAL3D+: Augments 12 rigid object classes of PASCAL VOC 2012 with 3D annotations.
• The PASCAL Visual Object Classes Challenges: Dataset and benchmarks for object class recognition.
• TME Motorway Dataset: 28 video sequences with vehicle annotations captured from VisLab's BRAiVE vehicle.
• LabelMe: Online annotation tool to build image databases for computer vision research.
• MIT Street Scenes: Street-side images with labels for 9 object categories (including cars, pedestrians, buildings, trees).
• Daimler Pedestrian Datasets: Datasets focusing on pedestrian detection for autonomous driving.
• Caltech Pedestrian Detection Benchmark: 10 hours of video with 350.000 annotated pedestrian bounding boxes.
• Robust Multi-Person Tracking from Mobile Platforms: Videos with annotated pedestrians captured from a stroller.

When using this dataset in your research, we will be happy if you cite us:
@INPROCEEDINGS{Geiger2012CVPR,
  author = {Andreas Geiger and Philip Lenz and Raquel Urtasun},
  title = {Are we ready for Autonomous Driving? The KITTI Vision Benchmark Suite},
  booktitle = {Conference on Computer Vision and Pattern Recognition (CVPR)},
  year = {2012}
}