Bird's Eye View Evaluation 2017


The bird's eye view benchmark consists of 7481 training images and 7518 test images as well as the corresponding point clouds, comprising a total of 80.256 labeled objects. For evaluation, we compute precision-recall curves. To rank the methods we compute average precision. 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.

We evaluate bird's eye view detection performance using the PASCAL criteria also used for 2D object detection. Far objects are thus filtered based on their bounding box height in the image plane. As only objects also appearing on the image plane are labeled, objects in don't car areas do not count as false positives. We note that the evaluation does not take care of ignoring detections that are not visible on the image plane — these detections might give rise to false positives. For cars we require a bounding box overlap of 70% in bird's eye view, while for pedestrians and cyclists we require an overlap of 50%. 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 2: On 08.10.2019, we have followed the suggestions of the Mapillary team in their paper Disentangling Monocular 3D Object Detection and use 40 recall positions instead of the 11 recall positions proposed in the original Pascal VOC benchmark. This results in a more fair comparison of the results, please check their paper. The last leaderboards right before this change can be found here: Object Detection Evaluation, 3D Object Detection Evaluation, Bird's Eye View Evaluation.
Important Policy Update: As more and more non-published work and re-implementations of existing work is submitted to KITTI, we have established a new policy: from now on, only submissions with significant novelty that are leading to a peer-reviewed paper in a conference or journal are allowed. Minor modifications of existing algorithms or student research projects are not allowed. Such work must be evaluated on a split of the training set. To ensure that our policy is adopted, new users must detail their status, describe their work and specify the targeted venue during registration. Furthermore, we will regularly delete all entries that are 6 months old but are still anonymous or do not have a paper associated with them. For conferences, 6 month is enough to determine if a paper has been accepted and to add the bibliography information. For longer review cycles, you need to resubmit your results.
Additional information used by the methods
  • Stereo: Method uses left and right (stereo) images
  • Flow: Method uses optical flow (2 temporally adjacent images)
  • Multiview: Method uses more than 2 temporally adjacent images
  • Laser Points: Method uses point clouds from Velodyne laser scanner
  • Additional training data: Use of additional data sources for training (see details)

Car


Method Setting Code Moderate Easy Hard Runtime Environment
1 GraR-Po 92.12 % 95.79 % 87.11 % 0.06 s 1 core @ 2.5 Ghz (Python + C/C++)
2 TED 92.05 % 95.44 % 87.30 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
3 LIVOX_Det
This method makes use of Velodyne laser scans.
 92.05 % 95.60 % 89.22 % n/a s 1 core @ 2.5 Ghz (Python + C/C++)
4 VPFNet 91.86 % 93.02 % 86.94 % 0.06 s 2 cores @ 2.5 Ghz (Python)
H. Zhu, J. Deng, Y. Zhang, J. Ji, Q. Mao, H. Li and Y. Zhang: VPFNet: Improving 3D Object Detection with Virtual Point based LiDAR and Stereo Data Fusion. 2021.
5 SFD code 91.85 % 95.64 % 86.83 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
X. Wu, L. Peng, H. Yang, L. Xie, C. Huang, C. Deng, H. Liu and D. Cai: Sparse Fuse Dense: Towards High Quality 3D Detection with Depth Completion. CVPR 2022.
6 SE-SSD
This method makes use of Velodyne laser scans.
 code 91.84 % 95.68 % 86.72 % 0.03 s 1 core @ 2.5 Ghz (Python + C/C++)
W. Zheng, W. Tang, L. Jiang and C. Fu: SE-SSD: Self-Ensembling Single-Stage Object Detector From Point Cloud. CVPR 2021.
7 GraR-Vo 91.72 % 95.27 % 86.51 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
8 PVT-SSD 91.63 % 95.23 % 86.43 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
9 CityBrainLab 91.62 % 94.78 % 86.68 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
10 SPANet 91.59 % 95.59 % 86.53 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
Y. Ye: SPANet: Spatial and Part-Aware Aggregation Network for 3D Object Detection. Pacific Rim International Conference on Artificial Intelligence 2021.
11 CasA 91.54 % 95.19 % 86.82 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
12 GraR-Pi 91.52 % 95.06 % 86.42 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
13 DGDNH 91.36 % 95.03 % 88.79 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
14 BADet code 91.32 % 95.23 % 86.48 % 0.14 s 1 core @ 2.5 Ghz (C/C++)
R. Qian, X. Lai and X. Li: BADet: Boundary-Aware 3D Object Detection from Point Clouds. Pattern Recognition 2022.
15 Anonymous 91.32 % 95.42 % 88.38 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
16 CasA++ 91.22 % 94.57 % 88.43 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
17 Anonymous 91.14 % 94.04 % 86.33 % n/a s 1 core @ 2.5 Ghz (C/C++)
18 SGFusion 91.11 % 94.76 % 86.27 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
19 Anonymous 91.04 % 94.76 % 86.31 % n/a s 1 core @ 2.5 Ghz (Python + C/C++)
20 SA-SSD code 91.03 % 95.03 % 85.96 % 0.04 s 1 core @ 2.5 Ghz (Python)
C. He, H. Zeng, J. Huang, X. Hua and L. Zhang: Structure Aware Single-stage 3D Object Detection from Point Cloud. CVPR 2020.
21 anonymous 90.90 % 92.96 % 86.34 % 0.09 s GPU @ 2.5 Ghz (Python)
22 MMLab PV-RCNN
This method makes use of Velodyne laser scans.
 code 90.65 % 94.98 % 86.14 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
S. Shi, C. Guo, L. Jiang, Z. Wang, J. Shi, X. Wang and H. Li: PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detection. CVPR 2020.
23 VueronNet code 90.56 % 94.67 % 85.31 % 0.06 s 1 core @ 2.0 Ghz (Python)
24 ST-RCNN
This method makes use of Velodyne laser scans.
 90.53 % 94.58 % 86.08 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
25 VPFNet code 90.52 % 93.94 % 86.25 % 0.2 s 1 core @ 2.5 Ghz (C/C++)
C. Wang, H. Chen and L. Fu: VPFNet: Voxel-Pixel Fusion Network for Multi-class 3D Object Detection. 2021.
26 PDV code 90.48 % 94.56 % 86.23 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
J. Hu, T. Kuai and S. Waslander: Point Density-Aware Voxels for LiDAR 3D Object Detection. CVPR 2022.
27 VCRCNN 90.42 % 94.55 % 86.11 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
28 M3DeTR code 90.37 % 94.41 % 85.98 % n/a s GPU @ 1.0 Ghz (Python)
T. Guan, J. Wang, S. Lan, R. Chandra, Z. Wu, L. Davis and D. Manocha: M3DeTR: Multi-representation, Multi- scale, Mutual-relation 3D Object Detection with Transformers. 2021.
29 TBD 90.37 % 93.82 % 87.72 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
30 DDet 90.34 % 94.16 % 86.01 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
31 VoTr-TSD code 90.34 % 94.03 % 86.14 % 0.07 s 1 core @ 2.5 Ghz (C/C++)
J. Mao, Y. Xue, M. Niu, H. Bai, J. Feng, X. Liang, H. Xu and C. Xu: Voxel Transformer for 3D Object Detection. ICCV 2021.
32 DSA-PV-RCNN
This method makes use of Velodyne laser scans.
 code 90.13 % 92.42 % 85.93 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
P. Bhattacharyya, C. Huang and K. Czarnecki: SA-Det3D: Self-Attention Based Context-Aware 3D Object Detection. 2021.
33 XView 90.12 % 92.27 % 85.94 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
L. Xie, G. Xu, D. Cai and X. He: X-view: Non-egocentric Multi-View 3D Object Detector. 2021.
34 GraR-VoI 90.10 % 95.69 % 86.85 % 0.07 s 1 core @ 2.5 Ghz (Python + C/C++)
35 CAT-Det 90.07 % 92.59 % 85.82 % 0.3 s GPU @ 2.5 Ghz (Python + C/C++)
Y. Zhang, J. Chen and D. Huang: CAT-Det: Contrastively Augmented Transformer for Multi-modal 3D Object Detection. CVPR 2022.
36 IKT3D
This method makes use of Velodyne laser scans.
 90.06 % 92.14 % 85.74 % 0.05 s 1 core @ 2.5 Ghz (Python)
37 FPV-SSD 89.93 % 91.45 % 85.21 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
38 SVGA-Net 89.88 % 92.07 % 85.59 % 0.03s 1 core @ 2.5 Ghz (Python + C/C++)
Q. He, Z. Wang, H. Zeng, Y. Zeng and Y. Liu: SVGA-Net: Sparse Voxel-Graph Attention Network for 3D Object Detection from Point Clouds. AAAI 2022.
39 EBM3DOD code 89.86 % 95.64 % 84.56 % 0.12 s 1 core @ 2.5 Ghz (Python)
F. Gustafsson, M. Danelljan and T. Schön: Accurate 3D Object Detection using Energy- Based Models. arXiv preprint arXiv:2012.04634 2020.
40 CIA-SSD
This method makes use of Velodyne laser scans.
 code 89.84 % 93.74 % 82.39 % 0.03 s 1 core @ 2.5 Ghz (Python + C/C++)
W. Zheng, W. Tang, S. Chen, L. Jiang and C. Fu: CIA-SSD: Confident IoU-Aware Single-Stage Object Detector From Point Cloud. AAAI 2021.
41 CLOCs_PVCas code 89.80 % 93.05 % 86.57 % 0.1 s 1 core @ 2.5 Ghz (Python)
S. Pang, D. Morris and H. Radha: CLOCs: Camera-LiDAR Object Candidates Fusion for 3D Object Detection . 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2020.
42 PE-RCVN 89.79 % 95.55 % 84.78 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
43 Anonymous 89.76 % 95.41 % 86.42 % 0.03 s 1 core @ 2.5 Ghz (Python + C/C++)
44 GLENet-VR 89.76 % 93.48 % 84.89 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
45 EBM3DOD baseline code 89.63 % 95.44 % 84.34 % 0.05 s 1 core @ 2.5 Ghz (Python)
F. Gustafsson, M. Danelljan and T. Schön: Accurate 3D Object Detection using Energy- Based Models. arXiv preprint arXiv:2012.04634 2020.
46 HCPVF 89.62 % 93.20 % 86.72 % 0.07 s 1 core @ 2.5 Ghz (Python + C/C++)
47 DSASNet 89.59 % 93.41 % 84.81 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
48 3SNet 89.58 % 93.26 % 84.80 % 0.07 s GPU @ 2.5 Ghz (Python)
49 CAD 89.57 % 93.03 % 84.71 % 0.1 s GPU @ 2.5 Ghz (Python + C/C++)
50 3D-CVF at SPA
This method makes use of Velodyne laser scans.
 89.56 % 93.52 % 82.45 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
J. Yoo, Y. Kim, J. Kim and J. Choi: 3D-CVF: Generating Joint Camera and LiDAR Features Using Cross-View Spatial Feature Fusion for 3D Object Detection. ECCV 2020.
51 ImpDet 89.55 % 92.74 % 84.41 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
52 Struc info fusion II 89.54 % 95.26 % 82.31 % 0.05 s GPU @ 2.5 Ghz (Python)
P. An, J. Liang, J. Ma, K. Yu and B. Fang: Struc info fusion. Submitted to CVIU 2021.
53 KpNet 89.53 % 93.34 % 81.95 % 0.42 s 1 core @ 2.5 Ghz (C/C++)
54 SASA
This method makes use of Velodyne laser scans.
 code 89.51 % 92.87 % 86.35 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
C. Chen, Z. Chen, J. Zhang and D. Tao: SASA: Semantics-Augmented Set Abstraction for Point-based 3D Object Detection. arXiv preprint arXiv:2201.01976 2022.
55 Fast-CLOCs 89.49 % 93.03 % 86.40 % 0.1 s GPU @ 2.5 Ghz (Python)
S. Pang, D. Morris and H. Radha: Fast-CLOCs: Fast Camera-LiDAR Object Candidates Fusion for 3D Object Detection. Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) 2022.
56 KpNet 89.49 % 93.29 % 81.92 % 42 s 1 core @ 2.5 Ghz (C/C++)
57 IA-SSD (single) code 89.48 % 93.14 % 84.42 % 0.013 s 1 core @ 2.5 Ghz (C/C++)
Y. Zhang, Q. Hu, G. Xu, Y. Ma, J. Wan and Y. Guo: Not All Points Are Equal: Learning Highly Efficient Point-based Detectors for 3D LiDAR Point Clouds. CVPR 2022.
58 CLOCs code 89.48 % 92.91 % 86.42 % 0.1 s 1 core @ 2.5 Ghz (Python)
S. Pang, D. Morris and H. Radha: CLOCs: Camera-LiDAR Object Candidates Fusion for 3D Object Detection . 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2020.
59 SA3DNet
This method uses stereo information.
This method makes use of Velodyne laser scans.
 89.46 % 93.11 % 84.60 % 0.05 s GPU @ 2.5 Ghz (Python)
60 DVF-V 89.42 % 93.12 % 86.50 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
A. Mahmoud, J. Hu and S. Waslander: Dense Voxel Fusion for 3D Object Detection. 2022.
61 Struc info fusion I 89.38 % 94.91 % 84.29 % 0.05 s 1 core @ 2.5 Ghz (Python)
P. An, J. Liang, J. Ma, K. Yu and B. Fang: Struc info fusion. Submitted to CVIU 2021.
62 JPVNet 89.36 % 92.78 % 84.37 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
63 BtcDet
This method makes use of Velodyne laser scans.
 code 89.34 % 92.81 % 84.55 % 0.09 s GPU @ 2.5 Ghz (Python + C/C++)
Q. Xu, Y. Zhong and U. Neumann: Behind the Curtain: Learning Occluded Shapes for 3D Object Detection. Proceedings of the AAAI Conference on Artificial Intelligence 2022.
64 IA-SSD (multi) code 89.33 % 92.79 % 84.35 % 0.014 s 1 core @ 2.5 Ghz (C/C++)
Y. Zhang, Q. Hu, G. Xu, Y. Ma, J. Wan and Y. Guo: Not All Points Are Equal: Learning Highly Efficient Point-based Detectors for 3D LiDAR Point Clouds. CVPR 2022.
65 Anonymous 89.27 % 92.79 % 86.53 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
66 TBD 89.24 % 92.59 % 85.99 % 0.1 s 1 core @ 2.5 Ghz (Python)
67 ATT_SSD 89.22 % 92.82 % 85.90 % 0.01 s 1 core @ 2.5 Ghz (Python)
68 TBD code 89.21 % 92.88 % 85.87 % 0.01 s 1 core @ 2.5 Ghz (C/C++)
69 DVF-PV 89.20 % 93.08 % 86.28 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
A. Mahmoud, J. Hu and S. Waslander: Dense Voxel Fusion for 3D Object Detection. 2022.
70 TF3D
This method makes use of Velodyne laser scans.
 89.19 % 93.10 % 84.41 % 0.1 s 2 cores @ 3.0 Ghz (Python)
71 STD code 89.19 % 94.74 % 86.42 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Yang, Y. Sun, S. Liu, X. Shen and J. Jia: STD: Sparse-to-Dense 3D Object Detector for Point Cloud. ICCV 2019.
72 FS-Net
This method makes use of Velodyne laser scans.
 89.18 % 92.88 % 84.39 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
73 Point-GNN
This method makes use of Velodyne laser scans.
 code 89.17 % 93.11 % 83.90 % 0.6 s GPU @ 2.5 Ghz (Python)
W. Shi and R. Rajkumar: Point-GNN: Graph Neural Network for 3D Object Detection in a Point Cloud. CVPR 2020.
74 HMFI code 89.17 % 93.04 % 86.37 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
75 SSL-PointGNN code 89.16 % 92.92 % 83.99 % 0.56 s GPU @ 1.5 Ghz (Python)
E. Erçelik, E. Yurtsever, M. Liu, Z. Yang, H. Zhang, P. Topçam, M. Listl, Y. Çaylı and A. Knoll: 3D Object Detection with a Self-supervised Lidar Scene Flow Backbone. arXiv preprint arXiv:2205.00705 2022.
76 Anonymous 89.15 % 92.47 % 85.97 % 0.06 s 1 core @ 2.5 Ghz (Python + C/C++)
77 SGNet 89.14 % 93.04 % 86.54 % 0.09 s GPU @ 2.5 Ghz (Python)
78 DGCN 89.14 % 92.62 % 83.90 % 0.1 s GPU @ 2.5 Ghz (Python)
79 USVLab BSAODet 89.13 % 92.92 % 86.41 % 0.04 s GPU @ 2.5 Ghz (Python + C/C++)
80 SPG_mini
This method makes use of Velodyne laser scans.
 code 89.12 % 92.80 % 86.27 % 0.09 s GPU @ 2.5 Ghz (Python)
Q. Xu, Y. Zhou, W. Wang, C. Qi and D. Anguelov: SPG: Unsupervised Domain Adaptation for 3D Object Detection via Semantic Point Generation. Proceedings of the IEEE conference on computer vision and pattern recognition (ICCV) 2021.
81 ITCA-SSD code 89.12 % 93.19 % 83.99 % 0.05 s 1 core @ 2.5 Ghz (Python)
82 EQ-PVRCNN code 89.09 % 94.55 % 86.42 % 0.2 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Yang, L. Jiang, Y. Sun, B. Schiele and J. Jia: A Unified Query-based Paradigm for Point Cloud Understanding. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2022.
83 SPT 89.09 % 94.87 % 84.38 % 0.1 s GPU @ 2.5 Ghz (Python)
84 MSADet 89.08 % 92.76 % 85.66 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
85 VoxSeT code 89.07 % 92.70 % 86.29 % 33 ms 1 core @ 2.5 Ghz (C/C++)
C. He, R. Li, S. Li and L. Zhang: Voxel Set Transformer: A Set-to-Set Approach to 3D Object Detection from Point Clouds. CVPR 2022.
86 3DSSD code 89.02 % 92.66 % 85.86 % 0.04 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Yang, Y. Sun, S. Liu and J. Jia: 3DSSD: Point-based 3D Single Stage Object Detector. CVPR 2020.
87 EPNet++ 89.00 % 95.41 % 85.73 % 0.1 s GPU @ 2.5 Ghz (Python)
Z. Liu, H. tengteng, B. Li, X. Chen, X. Wang and X. Bai: EPNet++: Cascade Bi-directional Fusion for Multi-Modal 3D Object Detection. arXiv preprint arXiv:2112.11088 2021.
88 Focals Conv code 89.00 % 92.67 % 86.33 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
Y. Chen, Y. Li, X. Zhang, J. Sun and J. Jia: Focal Sparse Convolutional Networks for 3D Object Detection. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2022.
89 SECOND 88.98 % 92.01 % 83.67 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
90 LGNet 88.98 % 92.83 % 86.26 % 0.07 s 1 core @ 2.5 Ghz (C/C++)
91 ISE-RCNN 88.97 % 92.86 % 86.28 % 0.09 s 1 core @ 2.5 Ghz (Python + C/C++)
92 PTA-RCNN 88.94 % 92.32 % 85.63 % 0.08 s 1 core @ 2.5 Ghz (Python)
93 GV-RCNN code 88.94 % 94.52 % 86.24 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
94 TBD 88.94 % 92.03 % 86.35 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
95 SPNet code 88.92 % 92.29 % 86.16 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
96 Sem-Aug v1 code 88.92 % 92.59 % 84.29 % 0.04 s GPU @ 3.5 Ghz (Python)
97 H^23D R-CNN code 88.87 % 92.85 % 86.07 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
J. Deng, W. Zhou, Y. Zhang and H. Li: From Multi-View to Hollow-3D: Hallucinated Hollow-3D R-CNN for 3D Object Detection. IEEE Transactions on Circuits and Systems for Video Technology 2021.
98 Pyramid R-CNN 88.84 % 92.19 % 86.21 % 0.07 s 1 core @ 2.5 Ghz (C/C++)
J. Mao, M. Niu, H. Bai, X. Liang, H. Xu and C. Xu: Pyramid R-CNN: Towards Better Performance and Adaptability for 3D Object Detection. ICCV 2021.
99 CityBrainLab-CT3D code 88.83 % 92.36 % 84.07 % 0.07 s 1 core @ 2.5 Ghz (Python + C/C++)
H. Sheng, S. Cai, Y. Liu, B. Deng, J. Huang, X. Hua and M. Zhao: Improving 3D Object Detection with Channel- wise Transformer. ICCV 2021.
100 Voxel R-CNN code 88.83 % 94.85 % 86.13 % 0.04 s GPU @ 3.0 Ghz (C/C++)
J. Deng, S. Shi, P. Li, W. Zhou, Y. Zhang and H. Li: Voxel R-CNN: Towards High Performance Voxel-based 3D Object Detection . AAAI 2021.
101 HVNet 88.82 % 92.83 % 83.38 % 0.03 s GPU @ 2.0 Ghz (Python)
M. Ye, S. Xu and T. Cao: HVNet: Hybrid Voxel Network for LiDAR Based 3D Object Detection. CVPR 2020.
102 GLENet 88.81 % 92.22 % 84.13 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
103 AGS-SSD[la] 88.80 % 92.61 % 85.44 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
104 FV2P v2 88.80 % 92.22 % 84.24 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
105 mbdf-netv1 code 88.77 % 94.45 % 83.90 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
106 SRIF-RCNN 88.77 % 92.10 % 86.06 % 0.0947 s 1 core @ 2.5 Ghz (C/C++)
X. Li and D. Kong: SRIF-RCNN: Sparsely Represented Inputs Fusion of Different Sensors for 3D Object Detection. Applied Intelligence 2022.
107 Anonymous 88.75 % 92.09 % 84.06 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
108 PV-RCNN++ code 88.74 % 92.66 % 85.97 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
109 SPG
This method makes use of Velodyne laser scans.
 code 88.70 % 94.33 % 85.98 % 0.09 s 1 core @ 2.5 Ghz (Python + C/C++)
Q. Xu, Y. Zhou, W. Wang, C. Qi and D. Anguelov: SPG: Unsupervised Domain Adaptation for 3D Object Detection via Semantic Point Generation. Proceedings of the IEEE conference on computer vision and pattern recognition (ICCV) 2021.
110 MVMM code 88.70 % 92.17 % 85.47 % 0.04 s GPU @ 2.5 Ghz (Python + C/C++)
111 ISE-RCNN-PV 88.69 % 92.31 % 86.10 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
112 DCAN-Second code 88.68 % 92.76 % 85.32 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
113 SIENet code 88.65 % 92.38 % 86.03 % 0.08 s 1 core @ 2.5 Ghz (Python)
Z. Li, Y. Yao, Z. Quan, W. Yang and J. Xie: SIENet: Spatial Information Enhancement Network for 3D Object Detection from Point Cloud. 2021.
114 P2V-RCNN 88.63 % 92.72 % 86.14 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
J. Li, S. Luo, Z. Zhu, H. Dai, A. Krylov, Y. Ding and L. Shao: P2V-RCNN: Point to Voxel Feature Learning for 3D Object Detection from Point Clouds. IEEE Access 2021.
115 FromVoxelToPoint code 88.61 % 92.23 % 86.11 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
J. Li, H. Dai, L. Shao and Y. Ding: From Voxel to Point: IoU-guided 3D Object Detection for Point Cloud with Voxel-to- Point Decoder. MM '21: The 29th ACM International Conference on Multimedia (ACM MM) 2021.
116 RangeIoUDet
This method makes use of Velodyne laser scans.
 88.59 % 92.28 % 85.83 % 0.02 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Liang, Z. Zhang, M. Zhang, X. Zhao and S. Pu: RangeIoUDet: Range Image Based Real-Time 3D Object Detector Optimized by Intersection Over Union. CVPR 2021.
117 WGVRF 88.56 % 92.45 % 85.69 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
118 DCCA 88.55 % 92.29 % 85.85 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
119 GVNet-V2 88.54 % 92.26 % 85.71 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
120 EPNet code 88.47 % 94.22 % 83.69 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
T. Huang, Z. Liu, X. Chen and X. Bai: EPNet: Enhancing Point Features with Image Semantics for 3D Object Detection. ECCV 2020.
121 GT3D 88.46 % 92.22 % 83.81 % 0.1 s 1 core @ 2.5 Ghz (Python)
122 CenterNet3D 88.46 % 91.80 % 83.62 % 0.04 s GPU @ 1.5 Ghz (Python)
G. Wang, B. Tian, Y. Ai, T. Xu, L. Chen and D. Cao: CenterNet3D:An Anchor free Object Detector for Autonomous Driving. 2020.
123 GVNet code 88.43 % 92.19 % 85.63 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
124 USVLab BSAODet (S) 88.42 % 92.19 % 85.55 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
125 DGT-Det3D code 88.41 % 92.57 % 85.50 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
126 Semantical PVRCNN 88.41 % 92.71 % 85.86 % 0.07 s 1 core @ 2.5 Ghz (C/C++)
127 RangeRCNN
This method makes use of Velodyne laser scans.
 88.40 % 92.15 % 85.74 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Liang, M. Zhang, Z. Zhang, X. Zhao and S. Pu: RangeRCNN: Towards Fast and Accurate 3D Object Detection with Range Image Representation. arXiv preprint arXiv:2009.00206 2020.
128 Patches
This method makes use of Velodyne laser scans.
 88.39 % 92.72 % 83.19 % 0.15 s GPU @ 2.0 Ghz
J. Lehner, A. Mitterecker, T. Adler, M. Hofmarcher, B. Nessler and S. Hochreiter: Patch Refinement: Localized 3D Object Detection. arXiv preprint arXiv:1910.04093 2019.
129 3D IoU-Net 88.38 % 94.76 % 81.93 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
J. Li, S. Luo, Z. Zhu, H. Dai, S. Krylov, Y. Ding and L. Shao: 3D IoU-Net: IoU Guided 3D Object Detector for Point Clouds. arXiv preprint arXiv:2004.04962 2020.
130 StructuralIF 88.38 % 91.78 % 85.67 % 0.02 s 8 cores @ 2.5 Ghz (Python)
J. Pei An: Deep structural information fusion for 3D object detection on LiDAR-camera system. Accepted in CVIU 2021.
131 CSVoxel-RCNN 88.37 % 92.07 % 85.51 % 0.03 s GPU @ 1.0 Ghz (Python)
132 NV-RCNN 88.36 % 91.41 % 85.72 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
133 DKDet 88.32 % 92.21 % 85.46 % 0.03 s GPU @ 2.5 Ghz (Python + C/C++)
134 CenterFuse 88.31 % 91.54 % 83.39 % 0.059 sec/frame 2 x V100
135 SARFE 88.28 % 92.35 % 85.50 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
136 FusionDetv2-v4 88.27 % 92.05 % 85.38 % 0.06 s 1 core @ 2.5 Ghz (Python + C/C++)
137 TBD 88.26 % 91.44 % 85.44 % 0.06 s GPU @ 2.5 Ghz (Python)
138 KPP3D code 88.25 % 93.93 % 83.26 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
139 CLOCs_SecCas 88.23 % 91.16 % 82.63 % 0.1 s 1 core @ 2.5 Ghz (Python)
S. Pang, D. Morris and H. Radha: CLOCs: Camera-LiDAR Object Candidates Fusion for 3D Object Detection. 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2020.
140 SPVB-SSD 88.23 % 91.82 % 85.46 % 0.04 s GPU @ 2.5 Ghz (Python + C/C++)
141 UberATG-MMF
This method makes use of Velodyne laser scans.
 88.21 % 93.67 % 81.99 % 0.08 s GPU @ 2.5 Ghz (Python)
M. Liang*, B. Yang*, Y. Chen, R. Hu and R. Urtasun: Multi-Task Multi-Sensor Fusion for 3D Object Detection. CVPR 2019.
142 Patches - EMP
This method makes use of Velodyne laser scans.
 88.17 % 94.49 % 84.75 % 0.5 s GPU @ 2.5 Ghz (Python)
J. Lehner, A. Mitterecker, T. Adler, M. Hofmarcher, B. Nessler and S. Hochreiter: Patch Refinement: Localized 3D Object Detection. arXiv preprint arXiv:1910.04093 2019.
143 SRDL 88.17 % 92.01 % 85.43 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
ERROR: Wrong syntax in BIBTEX file.
144 CF-cd-io-tv 88.16 % 91.32 % 83.26 % 1 s 1 core @ 2.5 Ghz (C/C++)
145 FusionDetv1 88.13 % 91.91 % 85.40 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
146 PSA-Det3D 88.13 % 92.08 % 85.35 % 0.1 s GPU @ 2.5 Ghz (Python)
147 PointPainting
This method makes use of Velodyne laser scans.
 88.11 % 92.45 % 83.36 % 0.4 s GPU @ 2.5 Ghz (Python + C/C++)
S. Vora, A. Lang, B. Helou and O. Beijbom: PointPainting: Sequential Fusion for 3D Object Detection. CVPR 2020.
148 SERCNN
This method makes use of Velodyne laser scans.
 88.10 % 94.11 % 83.43 % 0.1 s 1 core @ 2.5 Ghz (Python)
D. Zhou, J. Fang, X. Song, L. Liu, J. Yin, Y. Dai, H. Li and R. Yang: Joint 3D Instance Segmentation and Object Detection for Autonomous Driving. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition 2020.
149 Associate-3Ddet code 88.09 % 91.40 % 82.96 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
L. Du, X. Ye, X. Tan, J. Feng, Z. Xu, E. Ding and S. Wen: Associate-3Ddet: Perceptual-to-Conceptual Association for 3D Point Cloud Object Detection. The IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2020.
150 HotSpotNet 88.09 % 94.06 % 83.24 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
Q. Chen, L. Sun, Z. Wang, K. Jia and A. Yuille: object as hotspots. Proceedings of the European Conference on Computer Vision (ECCV) 2020.
151 Faraway-Frustum
This method makes use of Velodyne laser scans.
 code 88.08 % 91.90 % 85.35 % 0.1 s GPU @ 2.5 Ghz (Python)
H. Zhang, D. Yang, E. Yurtsever, K. Redmill and U. Ozguner: Faraway-frustum: Dealing with lidar sparsity for 3D object detection using fusion. 2021 IEEE International Intelligent Transportation Systems Conference (ITSC) 2021.
152 NV2P-RCNN 88.08 % 93.44 % 85.32 % 0.1 s GPU @ 2.5 Ghz (Python)
153 VPN 88.06 % 90.94 % 83.24 % 0.06 s 1 core @ 2.5 Ghz (Python + C/C++)
154 TBD 88.04 % 91.31 % 84.79 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
155 SC-Voxel-RCNN 88.02 % 91.45 % 85.22 % 0.12 s GPU @ 1.0 Ghz (Python)
156 CZY 88.00 % 91.85 % 85.22 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
157 UberATG-HDNET
This method makes use of Velodyne laser scans.
 87.98 % 93.13 % 81.23 % 0.05 s GPU @ 2.5 Ghz (Python)
B. Yang, M. Liang and R. Urtasun: HDNET: Exploiting HD Maps for 3D Object Detection. 2nd Conference on Robot Learning (CoRL) 2018.
158 TCDVF 87.94 % 91.21 % 84.66 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
159 DGT-Det3D 87.88 % 91.70 % 85.14 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
160 FusionDetv2-v5 87.86 % 91.92 % 83.07 % 0.05 s 1 core @ 2.5 Ghz (Java + C/C++)
161 Fast Point R-CNN
This method makes use of Velodyne laser scans.
 87.84 % 90.87 % 80.52 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
Y. Chen, S. Liu, X. Shen and J. Jia: Fast Point R-CNN. Proceedings of the IEEE international conference on computer vision (ICCV) 2019.
162 CSNet 87.84 % 92.23 % 82.93 % 0.1 s 1 core @ 2.5 Ghz (Python)
163 CF-ctdep-tv-ta 87.81 % 90.73 % 84.97 % 1 s 1 core @ 2.5 Ghz (C/C++)
164 MMLab-PartA^2
This method makes use of Velodyne laser scans.
 code 87.79 % 91.70 % 84.61 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
S. Shi, Z. Wang, J. Shi, X. Wang and H. Li: From Points to Parts: 3D Object Detection from Point Cloud with Part-aware and Part-aggregation Network. IEEE Transactions on Pattern Analysis and Machine Intelligence 2020.
165 cp-tv-kp-io-sc 87.78 % 90.98 % 84.04 % 1 s 1 core @ 2.5 Ghz (C/C++)
166 SIF 87.76 % 91.44 % 85.15 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
P. An: SIF. Submitted to CVIU 2021.
167 MVAF-Net code 87.73 % 91.95 % 85.00 % 0.06 s 1 core @ 2.5 Ghz (Python + C/C++)
G. Wang, B. Tian, Y. Zhang, L. Chen, D. Cao and J. Wu: Multi-View Adaptive Fusion Network for 3D Object Detection. arXiv preprint arXiv:2011.00652 2020.
168 Reprod-Two-Branch 87.69 % 90.69 % 84.72 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
169 DKAnet 87.68 % 91.07 % 84.03 % 0.05 s 1 core @ 2.0 Ghz (Python)
170 DVFENet 87.68 % 90.93 % 84.60 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
Y. He, G. Xia, Y. Luo, L. Su, Z. Zhang, W. Li and P. Wang: DVFENet: Dual-branch Voxel Feature Extraction Network for 3D Object Detection. Neurocomputing 2021.
171 S-AT GCN 87.68 % 90.85 % 84.20 % 0.02 s GPU @ 2.0 Ghz (Python)
L. Wang, C. Wang, X. Zhang, T. Lan and J. Li: S-AT GCN: Spatial-Attention Graph Convolution Network based Feature Enhancement for 3D Object Detection. CoRR 2021.
172 TBD 87.67 % 91.02 % 82.42 % 0.1 s 1 core @ 2.5 Ghz (Python)
173 CFF-tv-v2 87.67 % 90.70 % 84.58 % 1 s 1 core @ 2.5 Ghz (C/C++)
174 CFF-ep25 87.66 % 90.60 % 84.71 % 1 s 1 core @ 2.5 Ghz (C/C++)
175 TBD 87.62 % 90.86 % 82.29 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
176 CF-base-tv 87.60 % 90.28 % 84.52 % 1 s 1 core @ 2.5 Ghz (C/C++)
177 AutoAlign 87.60 % 91.72 % 84.44 % 0.1 s 1 core @ 2.5 Ghz (Python)
178 KeyFuse2B 87.59 % 90.70 % 84.58 % 0.01 s 1 core @ 2.5 Ghz (C/C++)
179 MODet
This method makes use of Velodyne laser scans.
 87.56 % 90.80 % 82.69 % 0.05 s GTX1080Ti
Y. Zhang, Z. Xiang, C. Qiao and S. Chen: Accurate and Real-Time Object Detection Based on Bird's Eye View on 3D Point Clouds. 2019 International Conference on 3D Vision (3DV) 2019.
180 CFF-tv 87.55 % 90.56 % 84.59 % 1 s 1 core @ 2.5 Ghz (C/C++)
181 cff-tv-v2-ep25 87.55 % 90.26 % 84.53 % 1 s 1 core @ 2.5 Ghz (C/C++)
182 AB3DMOT
This method makes use of Velodyne laser scans.
This is an online method (no batch processing).
 code 87.53 % 91.99 % 81.03 % 0.0047s 1 core @ 2.5 Ghz (Python)
X. Weng and K. Kitani: A Baseline for 3D Multi-Object Tracking. arXiv:1907.03961 2019.
183 TBD 87.51 % 90.76 % 80.15 % 0.1 s 1 core @ 2.5 Ghz (Python)
184 DTFI 87.51 % 91.01 % 84.25 % 0.03 s 1 core @ 2.5 Ghz (Python)
185 CF-ctdep-tv 87.50 % 90.56 % 84.65 % 1 s 1 core @ 2.5 Ghz (C/C++)
186 PointRGCN 87.49 % 91.63 % 80.73 % 0.26 s GPU @ V100 (Python)
J. Zarzar, S. Giancola and B. Ghanem: PointRGCN: Graph Convolution Networks for 3D Vehicles Detection Refinement. ArXiv 2019.
187 Anonymous 87.48 % 90.98 % 84.22 % 1 1 core @ 2.5 Ghz (Python)
188 MGAF-3DSSD code 87.47 % 92.70 % 82.19 % 0.1 s 1 core @ 2.5 Ghz (Python)
J. Li, H. Dai, L. Shao and Y. Ding: Anchor-free 3D Single Stage Detector with Mask-Guided Attention for Point Cloud. MM '21: The 29th ACM International Conference on Multimedia (ACM MM) 2021.
189 PC-CNN-V2
This method makes use of Velodyne laser scans.
 87.40 % 91.19 % 79.35 % 0.5 s GPU @ 2.5 Ghz (Matlab + C/C++)
X. Du, M. Ang, S. Karaman and D. Rus: A General Pipeline for 3D Detection of Vehicles. 2018 IEEE International Conference on Robotics and Automation (ICRA) 2018.
190 HS3D code 87.40 % 91.97 % 82.85 % 0.12 s 1 core @ 2.5 Ghz (Python + C/C++)
191 PVTr 87.39 % 91.21 % 84.77 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
192 MMLab-PointRCNN
This method makes use of Velodyne laser scans.
 code 87.39 % 92.13 % 82.72 % 0.1 s GPU @ 2.5 Ghz (Python + C/C++)
S. Shi, X. Wang and H. Li: Pointrcnn: 3d object proposal generation and detection from point cloud. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2019.
193 Sem-Aug 87.37 % 93.35 % 82.43 % 0.08 s GPU @ 2.5 Ghz (Python)
194 MAFF-Net(DAF-Pillar) 87.34 % 90.79 % 77.66 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
Z. Zhang, Z. Liang, M. Zhang, X. Zhao, Y. Ming, T. Wenming and S. Pu: MAFF-Net: Filter False Positive for 3D Vehicle Detection with Multi-modal Adaptive Feature Fusion. arXiv preprint arXiv:2009.10945 2020.
195 KeyPoint-IoUHead 87.32 % 90.36 % 83.23 % 0.01 s 1 core @ 2.5 Ghz (C/C++)
196 ZMMPP 87.25 % 90.47 % 82.42 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
197 HRI-VoxelFPN 87.21 % 92.75 % 79.82 % 0.02 s GPU @ 2.5 Ghz (Python + C/C++)
H. Kuang, B. Wang, J. An, M. Zhang and Z. Zhang: Voxel-FPN:multi-scale voxel feature aggregation in 3D object detection from point clouds. sensors 2020.
198 epBRM
This method makes use of Velodyne laser scans.
 code 87.13 % 90.70 % 81.92 % 0.1 s GPU @ >3.5 Ghz (Python + C/C++)
K. Shin: Improving a Quality of 3D Object Detection by Spatial Transformation Mechanism. arXiv preprint arXiv:1910.04853 2019.
199 3D_att
This method makes use of Velodyne laser scans.
 87.09 % 93.14 % 81.92 % 0.17 s GPU @ 2.5 Ghz (Python)
200 Contrastive PP code 87.06 % 92.99 % 81.96 % 0.01 s 1 core @ 2.5 Ghz (Python)
201 DVF 87.05 % 92.76 % 84.13 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
202 T_PVRCNN 86.97 % 91.63 % 82.20 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
203 SARPNET 86.92 % 92.21 % 81.68 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
Y. Ye, H. Chen, C. Zhang, X. Hao and Z. Zhang: SARPNET: Shape Attention Regional Proposal Network for LiDAR-based 3D Object Detection. Neurocomputing 2019.
204 cff-tv-t 86.92 % 91.04 % 80.46 % 1 s 1 core @ 2.5 Ghz (C/C++)
205 CZY_3917 86.90 % 90.69 % 82.22 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
206 CF-base-train 86.88 % 90.03 % 83.16 % 1 s 1 core @ 2.5 Ghz (C/C++)
207 Self-Calib Conv 86.86 % 90.00 % 83.88 % 0.01 s 1 core @ 2.5 Ghz (C/C++)
208 T_PVRCNN_V2 86.85 % 91.54 % 81.82 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
209 ARPNET 86.81 % 90.06 % 79.41 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
Y. Ye, C. Zhang and X. Hao: ARPNET: attention region proposal network for 3D object detection. Science China Information Sciences 2019.
210 C-GCN 86.78 % 91.11 % 80.09 % 0.147 s GPU @ V100 (Python)
J. Zarzar, S. Giancola and B. Ghanem: PointRGCN: Graph Convolution Networks for 3D Vehicles Detection Refinement. ArXiv 2019.
211 IoU-2B 86.74 % 90.92 % 80.40 % 0.01 s 1 core @ 2.5 Ghz (C/C++)
212 cp-tv-kp 86.58 % 89.58 % 83.64 % 1 s 1 core @ 2.5 Ghz (C/C++)
213 PointPillars
This method makes use of Velodyne laser scans.
 code 86.56 % 90.07 % 82.81 % 16 ms 1080ti GPU and Intel i7 CPU
A. Lang, S. Vora, H. Caesar, L. Zhou, J. Yang and O. Beijbom: PointPillars: Fast Encoders for Object Detection from Point Clouds. CVPR 2019.
214 TANet code 86.54 % 91.58 % 81.19 % 0.035s GPU @ 2.5 Ghz (Python + C/C++)
Z. Liu, X. Zhao, T. Huang, R. Hu, Y. Zhou and X. Bai: TANet: Robust 3D Object Detection from Point Clouds with Triple Attention. AAAI 2020.
215 cp-tv 86.52 % 89.55 % 83.45 % 1 s 1 core @ 2.5 Ghz (C/C++)
216 SCNet
This method makes use of Velodyne laser scans.
 86.48 % 90.07 % 81.30 % 0.04 s GPU @ 3.0 Ghz (Python)
Z. Wang, H. Fu, L. Wang, L. Xiao and B. Dai: SCNet: Subdivision Coding Network for Object Detection Based on 3D Point Cloud. IEEE Access 2019.
217 CF-ctdep-train 86.46 % 89.57 % 82.03 % 1 s 1 core @ 2.5 Ghz (C/C++)
218 CSNet8306 code 86.44 % 92.57 % 81.36 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
219 SegVoxelNet 86.37 % 91.62 % 83.04 % 0.04 s 1 core @ 2.5 Ghz (Python)
H. Yi, S. Shi, M. Ding, J. Sun, K. Xu, H. Zhou, Z. Wang, S. Li and G. Wang: SegVoxelNet: Exploring Semantic Context and Depth-aware Features for 3D Vehicle Detection from Point Cloud. ICRA 2020.
220 Dune-DCF-e09 86.36 % 89.33 % 81.77 % 1 s 1 core @ 2.5 Ghz (C/C++)
221 Dune-DCF-e11 86.32 % 89.32 % 81.78 % 1 s 1 core @ 2.5 Ghz (C/C++)
222 PP-PCdet code 86.32 % 89.86 % 81.62 % 0.01 s 1 core @ 2.5 Ghz (Python)
223 3D IoU Loss
This method makes use of Velodyne laser scans.
 86.22 % 91.36 % 81.20 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
D. Zhou, J. Fang, X. Song, C. Guan, J. Yin, Y. Dai and R. Yang: IoU Loss for 2D/3D Object Detection. International Conference on 3D Vision (3DV) 2019.
224 Dune-DCF-e15 86.21 % 88.99 % 81.62 % 1 s 1 core @ 2.5 Ghz (C/C++)
225 TBD_BD code 86.12 % 91.00 % 81.66 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
226 CrazyTensor-CF 86.10 % 89.13 % 81.61 % 1 s 1 core @ 2.5 Ghz (C/C++)
227 City-CF-fixed 86.09 % 89.94 % 81.73 % 1 s 1 core @ 2.5 Ghz (C/C++)
228 R-GCN 86.05 % 91.91 % 81.05 % 0.16 s GPU @ 2.5 Ghz (Python)
J. Zarzar, S. Giancola and B. Ghanem: PointRGCN: Graph Convolution Networks for 3D Vehicles Detection Refinement. ArXiv 2019.
229 UberATG-PIXOR++
This method makes use of Velodyne laser scans.
 86.01 % 93.28 % 80.11 % 0.035 s GPU @ 2.5 Ghz (Python)
B. Yang, M. Liang and R. Urtasun: HDNET: Exploiting HD Maps for 3D Object Detection. 2nd Conference on Robot Learning (CoRL) 2018.
230 SSL_PP code 85.93 % 92.19 % 80.40 % 16ms GPU @ 1.5 Ghz (Python)
231 CSNet8299 code 85.91 % 91.64 % 80.95 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
232 Sem-Aug-PointRCNN++ 85.88 % 91.68 % 83.37 % 0.1 s 8 cores @ 3.0 Ghz (Python)
233 DASS 85.85 % 91.74 % 80.97 % 0.09 s 1 core @ 2.0 Ghz (Python)
O. Unal, L. Van Gool and D. Dai: Improving Point Cloud Semantic Segmentation by Learning 3D Object Detection. Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) 2021.
234 F-ConvNet
This method makes use of Velodyne laser scans.
 code 85.84 % 91.51 % 76.11 % 0.47 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Wang and K. Jia: Frustum ConvNet: Sliding Frustums to Aggregate Local Point-Wise Features for Amodal 3D Object Detection. IROS 2019.
235 City-CF 85.83 % 89.20 % 81.61 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
236 PI-RCNN 85.81 % 91.44 % 81.00 % 0.1 s 1 core @ 2.5 Ghz (Python)
L. Xie, C. Xiang, Z. Yu, G. Xu, Z. Yang, D. Cai and X. He: PI-RCNN: An Efficient Multi-sensor 3D Object Detector with Point-based Attentive Cont-conv Fusion Module. AAAI 2020 : The Thirty-Fourth AAAI Conference on Artificial Intelligence 2020.
237 LazyTorch-CP-Infer-O 85.74 % 89.19 % 81.35 % 1 s 1 core @ 2.5 Ghz (C/C++)
238 PointRGBNet 85.73 % 91.39 % 80.68 % 0.08 s 4 cores @ 2.5 Ghz (Python + C/C++)
P. Xie Desheng: Real-time Detection of 3D Objects Based on Multi-Sensor Information Fusion. Automotive Engineering 2022.
239 AFTD 85.63 % 90.61 % 82.28 % 1 s 1 core @ 2.5 Ghz (Python + C/C++)
240 LazyTorch-CP-Small-P 85.63 % 89.10 % 81.27 % 1 s 1 core @ 2.5 Ghz (C/C++)
241 CrazyTensor-CP 85.55 % 87.94 % 82.63 % 1 s 1 core @ 2.5 Ghz (Python)
242 Sem-Aug-PointRCNN code 85.50 % 89.75 % 83.13 % 0.1 s GPU @ 3.5 Ghz (C/C++)
243 variance_point 85.39 % 91.90 % 81.13 % 0.05 s 1 core @ 2.5 Ghz (Python)
244 UberATG-ContFuse
This method makes use of Velodyne laser scans.
 85.35 % 94.07 % 75.88 % 0.06 s GPU @ 2.5 Ghz (Python)
M. Liang, B. Yang, S. Wang and R. Urtasun: Deep Continuous Fusion for Multi-Sensor 3D Object Detection. ECCV 2018.
245 new_stereo 85.24 % 90.74 % 82.10 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
246 PSM_stereo 85.12 % 90.26 % 80.21 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
247 PFF3D
This method makes use of Velodyne laser scans.
 code 85.08 % 89.61 % 80.42 % 0.05 s GPU @ 3.0 Ghz (Python + C/C++)
L. Wen and K. Jo: Fast and Accurate 3D Object Detection for Lidar-Camera-Based Autonomous Vehicles Using One Shared Voxel-Based Backbone. IEEE Access 2021.
248 CenterPoint (pcdet) 85.05 % 88.47 % 81.19 % 0.051 sec/frame 2 x V100
249 AVOD
This method makes use of Velodyne laser scans.
 code 84.95 % 89.75 % 78.32 % 0.08 s Titan X (pascal)
J. Ku, M. Mozifian, J. Lee, A. Harakeh and S. Waslander: Joint 3D Proposal Generation and Object Detection from View Aggregation. IROS 2018.
250 WS3D
This method makes use of Velodyne laser scans.
 84.93 % 90.96 % 77.96 % 0.1 s GPU @ 2.5 Ghz (Python)
Q. Meng, W. Wang, T. Zhou, J. Shen, L. Van Gool and D. Dai: Weakly Supervised 3D Object Detection from Lidar Point Cloud. 2020.
251 AVOD-FPN
This method makes use of Velodyne laser scans.
 code 84.82 % 90.99 % 79.62 % 0.1 s Titan X (Pascal)
J. Ku, M. Mozifian, J. Lee, A. Harakeh and S. Waslander: Joint 3D Proposal Generation and Object Detection from View Aggregation. IROS 2018.
252 MF 84.72 % 88.58 % 78.17 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
253 F-PointNet
This method makes use of Velodyne laser scans.
 code 84.67 % 91.17 % 74.77 % 0.17 s GPU @ 3.0 Ghz (Python)
C. Qi, W. Liu, C. Wu, H. Su and L. Guibas: Frustum PointNets for 3D Object Detection from RGB-D Data. arXiv preprint arXiv:1711.08488 2017.
254 FusionDetv2-baseline 84.31 % 90.38 % 79.23 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
255 3DBN
This method makes use of Velodyne laser scans.
 83.94 % 89.66 % 76.50 % 0.13s 1080Ti (Python+C/C++)
X. Li, J. Guivant, N. Kwok and Y. Xu: 3D Backbone Network for 3D Object Detection. CoRR 2019.
256 MLOD
This method makes use of Velodyne laser scans.
 code 82.68 % 90.25 % 77.97 % 0.12 s GPU @ 1.5 Ghz (Python)
J. Deng and K. Czarnecki: MLOD: A multi-view 3D object detection based on robust feature fusion method. arXiv preprint arXiv:1909.04163 2019.
257 BirdNet+
This method makes use of Velodyne laser scans.
 code 81.85 % 87.43 % 75.36 % 0.11 s Titan Xp (PyTorch)
A. Barrera, J. Beltrán, C. Guindel, J. Iglesias and F. García: BirdNet+: Two-Stage 3D Object Detection in LiDAR through a Sparsity-Invariant Bird’s Eye View. IEEE Access 2021.
258 TBD 81.53 % 87.90 % 74.26 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
259 FD 81.47 % 88.34 % 75.07 % 0.01 s 1 core @ 2.5 Ghz (C/C++)
260 CZY 81.21 % 89.10 % 76.13 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
261 AEC3D 80.37 % 86.81 % 74.26 % 18 ms GPU @ 2.5 Ghz (Python)
262 DMF
This method uses stereo information.
 80.29 % 84.64 % 76.05 % 0.2 s 1 core @ 2.5 Ghz (Python + C/C++)
X. J. Chen and W. Xu: Disparity-Based Multiscale Fusion Network for Transportation Detection. IEEE Transactions on Intelligent Transportation Systems 2022.
263 UberATG-PIXOR
This method makes use of Velodyne laser scans.
 80.01 % 83.97 % 74.31 % 0.035 s TITAN Xp (Python)
B. Yang, W. Luo and R. Urtasun: PIXOR: Real-time 3D Object Detection from Point Clouds. CVPR 2018.
264 MV3D (LIDAR)
This method makes use of Velodyne laser scans.
 78.98 % 86.49 % 72.23 % 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.
265 DSGN++
This method uses stereo information.
 code 78.94 % 88.55 % 69.74 % 0.2 s GeForce RTX 2080Ti
Y. Chen, S. Huang, S. Liu, B. Yu and J. Jia: DSGN++: Exploiting Visual-Spatial Relation for Stereo-based 3D Detectors. arXiv preprint arXiv:2204.03039 2022.
266 MV3D
This method makes use of Velodyne laser scans.
 78.93 % 86.62 % 69.80 % 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.
267 StereoDistill 78.59 % 89.03 % 69.34 % 0.4 s 1 core @ 2.5 Ghz (Python)
268 Anonymous 77.40 % 90.76 % 70.00 % 0.5 s 1 core @ 2.5 Ghz (C/C++)
269 SD3DOD 76.96 % 86.82 % 70.05 % 0.04 s GPU @ 2.5 Ghz (Python)
270 MMLAB LIGA-Stereo
This method uses stereo information.
 code 76.78 % 88.15 % 67.40 % 0.4 s 1 core @ 2.5 Ghz (Python + C/C++)
X. Guo, S. Shi, X. Wang and H. Li: LIGA-Stereo: Learning LiDAR Geometry Aware Representations for Stereo-based 3D Detector. Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) 2021.
271 RCD 75.83 % 82.26 % 69.61 % 0.1 s GPU @ 2.5 Ghz (Python)
A. Bewley, P. Sun, T. Mensink, D. Anguelov and C. Sminchisescu: Range Conditioned Dilated Convolutions for Scale Invariant 3D Object Detection. Conference on Robot Learning (CoRL) 2020.
272 LaserNet 74.52 % 79.19 % 68.45 % 12 ms GPU @ 2.5 Ghz (C/C++)
G. Meyer, A. Laddha, E. Kee, C. Vallespi-Gonzalez and C. Wellington: LaserNet: An Efficient Probabilistic 3D Object Detector for Autonomous Driving. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2019.
273 PL++ (SDN+GDC)
This method uses stereo information.
This method makes use of Velodyne laser scans.
 code 73.80 % 84.61 % 65.59 % 0.6 s GPU @ 2.5 Ghz (C/C++)
Y. You, Y. Wang, W. Chao, D. Garg, G. Pleiss, B. Hariharan, M. Campbell and K. Weinberger: Pseudo-LiDAR++: Accurate Depth for 3D Object Detection in Autonomous Driving. International Conference on Learning Representations 2020.
274 SNVC
This method uses stereo information.
 code 73.61 % 86.88 % 64.49 % 1 s GPU @ 1.0 Ghz (Python)
S. Li, Z. Liu, Z. Shen and K. Cheng: Stereo Neural Vernier Caliper. Proceedings of the AAAI Conference on Artificial Intelligence 2022.
275 A3DODWTDA
This method makes use of Velodyne laser scans.
 code 73.26 % 79.58 % 62.77 % 0.08 s GPU @ 3.0 Ghz (Python)
F. Gustafsson and E. Linder-Norén: Automotive 3D Object Detection Without Target Domain Annotations. 2018.
276 Anonymous 71.23 % 86.67 % 64.08 % 0.5 s 1 core @ 2.5 Ghz (C/C++)
277 ppt 70.21 % 72.17 % 65.26 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
278 Complexer-YOLO
This method makes use of Velodyne laser scans.
 68.96 % 77.24 % 64.95 % 0.06 s GPU @ 3.5 Ghz (C/C++)
M. Simon, K. Amende, A. Kraus, J. Honer, T. Samann, H. Kaulbersch, S. Milz and H. Michael Gross: Complexer-YOLO: Real-Time 3D Object Detection and Tracking on Semantic Point Clouds. The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops 2019.
279 PS++ code 68.36 % 84.64 % 59.01 % PS++ s 1 core @ 2.5 Ghz (C/C++)
280 TopNet-Retina
This method makes use of Velodyne laser scans.
 68.16 % 80.16 % 63.43 % 52ms GeForce 1080Ti (tensorflow-gpu, v1.12)
S. Wirges, T. Fischer, C. Stiller and J. Frias: Object Detection and Classification in Occupancy Grid Maps Using Deep Convolutional Networks. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
281 CG-Stereo
This method uses stereo information.
 66.44 % 85.29 % 58.95 % 0.57 s GeForce RTX 2080 Ti
C. Li, J. Ku and S. Waslander: Confidence Guided Stereo 3D Object Detection with Split Depth Estimation. IROS 2020.
282 PLUME
This method uses stereo information.
 66.27 % 82.97 % 56.70 % 0.15 s GPU @ 2.5 Ghz (Python)
Y. Wang, B. Yang, R. Hu, M. Liang and R. Urtasun: PLUME: Efficient 3D Object Detection from Stereo Images. IROS 2021.
283 CDN
This method uses stereo information.
 code 66.24 % 83.32 % 57.65 % 0.6 s GPU @ 2.5 Ghz (Python)
D. Garg, Y. Wang, B. Hariharan, M. Campbell, K. Weinberger and W. Chao: Wasserstein Distances for Stereo Disparity Estimation. Advances in Neural Information Processing Systems (NeurIPS) 2020.
284 PS code 65.33 % 83.75 % 56.14 % PS s 1 core @ 2.5 Ghz (C/C++)
285 DSGN
This method uses stereo information.
 code 65.05 % 82.90 % 56.60 % 0.67 s NVIDIA Tesla V100
Y. Chen, S. Liu, X. Shen and J. Jia: DSGN: Deep Stereo Geometry Network for 3D Object Detection. CVPR 2020.
286 TopNet-DecayRate
This method makes use of Velodyne laser scans.
 64.60 % 79.74 % 58.04 % 92 ms NVIDIA GeForce 1080 Ti (tensorflow-gpu)
S. Wirges, T. Fischer, C. Stiller and J. Frias: Object Detection and Classification in Occupancy Grid Maps Using Deep Convolutional Networks. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
287 UPF_3D
This method uses stereo information.
 63.58 % 85.53 % 56.56 % 0.29 s 1 core @ 2.5 Ghz (Python)
288 BirdNet+ (legacy)
This method makes use of Velodyne laser scans.
 code 63.33 % 84.80 % 61.23 % 0.1 s Titan Xp (PyTorch)
A. Barrera, C. Guindel, J. Beltrán and F. García: BirdNet+: End-to-End 3D Object Detection in LiDAR Bird’s Eye View. 2020 IEEE 23rd International Conference on Intelligent Transportation Systems (ITSC) 2020.
289 3D FCN
This method makes use of Velodyne laser scans.
 61.67 % 70.62 % 55.61 % >5 s 1 core @ 2.5 Ghz (C/C++)
B. Li: 3D Fully Convolutional Network for Vehicle Detection in Point Cloud. IROS 2017.
290 CDN-PL++
This method uses stereo information.
 61.04 % 81.27 % 52.84 % 0.4 s GPU @ 2.5 Ghz (C/C++)
D. Garg, Y. Wang, B. Hariharan, M. Campbell, K. Weinberger and W. Chao: Wasserstein Distances for Stereo Disparity Estimation. Advances in Neural Information Processing Systems 2020.
291 BirdNet
This method makes use of Velodyne laser scans.
 59.83 % 84.17 % 57.35 % 0.11 s Titan Xp (Caffe)
J. Beltrán, C. Guindel, F. Moreno, D. Cruzado, F. García and A. Escalera: BirdNet: A 3D Object Detection Framework from LiDAR Information. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
292 TopNet-UncEst
This method makes use of Velodyne laser scans.
 59.67 % 72.05 % 51.67 % 0.09 s NVIDIA GeForce 1080 Ti (tensorflow-gpu)
S. Wirges, M. Braun, M. Lauer and C. Stiller: Capturing Object Detection Uncertainty in Multi-Layer Grid Maps. 2019.
293 RT3D-GMP
This method uses stereo information.
 59.00 % 69.14 % 45.49 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
H. Königshof and C. Stiller: Learning-Based Shape Estimation with Grid Map Patches for Realtime 3D Object Detection for Automated Driving. 2020 IEEE 23rd International Conference on Intelligent Transportation Systems (ITSC) 2020.
294 Disp R-CNN (velo)
This method uses stereo information.
 code 58.62 % 79.76 % 47.73 % 0.387 s GPU @ 2.5 Ghz (Python + C/C++)
J. Sun, L. Chen, Y. Xie, S. Zhang, Q. Jiang, X. Zhou and H. Bao: Disp R-CNN: Stereo 3D Object Detection via Shape Prior Guided Instance Disparity Estimation. CVPR 2020.
295 ESGN
This method uses stereo information.
 58.12 % 78.10 % 49.28 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
296 Pseudo-LiDAR++
This method uses stereo information.
 code 58.01 % 78.31 % 51.25 % 0.4 s GPU @ 2.5 Ghz (Python)
Y. You, Y. Wang, W. Chao, D. Garg, G. Pleiss, B. Hariharan, M. Campbell and K. Weinberger: Pseudo-LiDAR++: Accurate Depth for 3D Object Detection in Autonomous Driving. International Conference on Learning Representations 2020.
297 Disp R-CNN
This method uses stereo information.
 code 57.98 % 79.61 % 47.09 % 0.387 s GPU @ 2.5 Ghz (Python + C/C++)
J. Sun, L. Chen, Y. Xie, S. Zhang, Q. Jiang, X. Zhou and H. Bao: Disp R-CNN: Stereo 3D Object Detection via Shape Prior Guided Instance Disparity Estimation. CVPR 2020.
298 ZoomNet
This method uses stereo information.
 code 54.91 % 72.94 % 44.14 % 0.3 s 1 core @ 2.5 Ghz (C/C++)
L. Z. Xu: ZoomNet: Part-Aware Adaptive Zooming Neural Network for 3D Object Detection. Proceedings of the AAAI Conference on Artificial Intelligence 2020.
299 ART 54.23 % 75.05 % 48.19 % 20ms s 1 core @ 2.5 Ghz (C/C++)
300 VoxelJones code 53.96 % 66.21 % 47.66 % .18 s 1 core @ 2.5 Ghz (Python + C/C++)
M. Motro and J. Ghosh: Vehicular Multi-object Tracking with Persistent Detector Failures. arXiv preprint arXiv:1907.11306 2019.
301 TopNet-HighRes
This method makes use of Velodyne laser scans.
 53.05 % 67.84 % 46.99 % 101ms NVIDIA GeForce 1080 Ti (tensorflow-gpu)
S. Wirges, T. Fischer, C. Stiller and J. Frias: Object Detection and Classification in Occupancy Grid Maps Using Deep Convolutional Networks. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
302 OC Stereo
This method uses stereo information.
 code 51.47 % 68.89 % 42.97 % 0.35 s 1 core @ 2.5 Ghz (Python + C/C++)
A. Pon, J. Ku, C. Li and S. Waslander: Object-Centric Stereo Matching for 3D Object Detection. ICRA 2020.
303 YOLOStereo3D
This method uses stereo information.
 code 50.28 % 76.10 % 36.86 % 0.1 s GPU 1080Ti
Y. Liu, L. Wang and M. Liu: YOLOStereo3D: A Step Back to 2D for Efficient Stereo 3D Detection. 2021 International Conference on Robotics and Automation (ICRA) 2021.
304 RT3DStereo
This method uses stereo information.
 46.82 % 58.81 % 38.38 % 0.08 s GPU @ 2.5 Ghz (C/C++)
H. Königshof, N. Salscheider and C. Stiller: Realtime 3D Object Detection for Automated Driving Using Stereo Vision and Semantic Information. Proc. IEEE Intl. Conf. Intelligent Transportation Systems 2019.
305 Pseudo-Lidar
This method uses stereo information.
 code 45.00 % 67.30 % 38.40 % 0.4 s GPU @ 2.5 Ghz (Python + C/C++)
Y. Wang, W. Chao, D. Garg, B. Hariharan, M. Campbell and K. Weinberger: Pseudo-LiDAR From Visual Depth Estimation: Bridging the Gap in 3D Object Detection for Autonomous Driving. The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2019.
306 RT3D
This method makes use of Velodyne laser scans.
 44.00 % 56.44 % 42.34 % 0.09 s GPU @ 1.8Ghz
Y. Zeng, Y. Hu, S. Liu, J. Ye, Y. Han, X. Li and N. Sun: RT3D: Real-Time 3-D Vehicle Detection in LiDAR Point Cloud for Autonomous Driving. IEEE Robotics and Automation Letters 2018.
307 Stereo CenterNet
This method uses stereo information.
 42.12 % 62.97 % 35.37 % 0.04 s GPU @ 2.5 Ghz (Python)
Y. Shi, Y. Guo, Z. Mi and X. Li: Stereo CenterNet-based 3D object detection for autonomous driving. Neurocomputing 2022.
308 Stereo R-CNN
This method uses stereo information.
 code 41.31 % 61.92 % 33.42 % 0.3 s GPU @ 2.5 Ghz (Python)
P. Li, X. Chen and S. Shen: Stereo R-CNN based 3D Object Detection for Autonomous Driving. CVPR 2019.
309 SparseLiDAR_fusion 38.99 % 52.57 % 32.86 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
310 GCDR 37.34 % 50.85 % 30.51 % 0.28 s 1 core @ 2.5 Ghz (Python)
311 VMDet_Boost 33.13 % 46.17 % 28.80 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
312 StereoFENet
This method uses stereo information.
 32.96 % 49.29 % 25.90 % 0.15 s 1 core @ 3.5 Ghz (Python)
W. Bao, B. Xu and Z. Chen: MonoFENet: Monocular 3D Object Detection with Feature Enhancement Networks. IEEE Transactions on Image Processing 2019.
313 Anonymous 30.81 % 43.11 % 26.81 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
314 Digging_M3D 28.84 % 39.74 % 26.08 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
315 Mobile Stereo R-CNN
This method uses stereo information.
 28.78 % 44.51 % 22.30 % 1.8 s NVIDIA Jetson TX2
M. Hussein, M. Khalil and B. Abdullah: 3D Object Detection using Mobile Stereo R- CNN on Nvidia Jetson TX2. International Conference on Advanced Engineering, Technology and Applications (ICAETA) 2021.
316 VMDet 28.50 % 41.41 % 23.88 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
317 Anonymous 27.70 % 37.81 % 24.61 % 40 s 1 core @ 2.5 Ghz (C/C++)
318 SARM3D 26.81 % 34.17 % 23.68 % 0.03 s GPU @ 2.5 Ghz (Python)
319 CMKD* 25.82 % 38.98 % 22.80 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
320 MoGDE 25.60 % 38.38 % 22.91 % 0.03 s GPU @ 2.5 Ghz (Python)
321 LPCG-Monoflex 24.81 % 35.96 % 21.86 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
322 Anonymous 24.78 % 33.38 % 22.00 % 40 s 1 core @ 2.5 Ghz (C/C++)
323 DD3Dv2 code 24.67 % 35.70 % 21.73 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
324 Mix-Teaching 24.23 % 35.74 % 20.80 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
325 MonoInsight 24.23 % 34.85 % 20.87 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
326 anonymity 23.92 % 36.92 % 21.68 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
327 PS-fld code 23.76 % 32.64 % 20.64 % 0.25 s 1 core @ 2.5 Ghz (C/C++)
Y. Chen, H. Dai and Y. Ding: Pseudo-Stereo for Monocular 3D Object Detection in Autonomous Driving. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2022.
328 SCSTSV-MonoFlex 23.71 % 34.59 % 20.41 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
329 anonymity 23.61 % 36.80 % 21.31 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
330 MonoDDE 23.46 % 33.58 % 20.37 % 0.04 s 1 core @ 2.5 Ghz (Python)
Z. Li, Z. Qu, Y. Zhou, J. Liu, H. Wang and L. Jiang: Diversity Matters: Fully Exploiting Depth Clues for Reliable Monocular 3D Object Detection. CVPR 2022.
331 DD3D code 23.41 % 32.35 % 20.42 % n/a s 1 core @ 2.5 Ghz (C/C++)
D. Park, R. Ambrus, V. Guizilini, J. Li and A. Gaidon: Is Pseudo-Lidar needed for Monocular 3D Object detection?. IEEE/CVF International Conference on Computer Vision (ICCV) .
332 DID-M3D 22.76 % 32.95 % 19.83 % 0.04 s 1 core @ 2.5 Ghz (Python)
333 MonoDistill 22.59 % 31.87 % 19.72 % 0.04 s 1 core @ 2.5 Ghz (Python)
334 zongmuDistill 22.56 % 33.48 % 19.88 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
335 OPA-3D code 22.53 % 33.54 % 19.22 % 0.04 s 1 core @ 3.5 Ghz (Python)
336 Shape-Aware 22.13 % 32.55 % 18.94 % 0.05 s 1 core @ 2.5 Ghz (Python)
337 MonoCon code 22.10 % 31.12 % 19.00 % 0.02 s GPU @ 2.5 Ghz (Python)
T. Xianpeng Liu: Learning Auxiliary Monocular Contexts Helps Monocular 3D Object Detection. AAAI 2022.
338 Anonymous 22.05 % 31.75 % 19.44 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
339 gupnet_se 21.98 % 32.82 % 18.70 % 0.03s 1 core @ 2.5 Ghz (C/C++)
340 ZongmuMono3d code 21.78 % 33.18 % 18.71 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
341 MDNet 21.71 % 33.31 % 18.49 % 0.2 s 1 core @ 2.5 Ghz (C/C++)
342 Lite-FPN-GUPNet 21.53 % 31.68 % 18.38 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
343 DDS code 21.50 % 32.55 % 18.25 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
344 MonoDETR code 21.45 % 32.20 % 18.68 % 0.04 s 1 core @ 2.5 Ghz (Python)
R. Zhang, H. Qiu, T. Wang, X. Xu, Z. Guo, Y. Qiao, P. Gao and H. Li: MonoDETR: Depth-aware Transformer for Monocular 3D Object Detection. arXiv preprint arXiv:2203.13310 2022.
345 OBMO_GUPNet 21.41 % 30.81 % 18.37 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
346 M3DGAF 21.39 % 31.34 % 19.28 % 0.07 s 1 core @ 2.5 Ghz (Python)
347 mono3d code 21.39 % 32.17 % 18.47 % TBD TBD
348 SGM3D 21.37 % 31.49 % 18.43 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
Z. Zhou, L. Du, X. Ye, Z. Zou, X. Tan, E. Ding, L. Zhang, X. Xue and J. Feng: SGM3D: Stereo Guided Monocular 3D Object Detection. 2021.
349 monopd code 21.29 % 32.12 % 18.08 % 0.01 s 1 core @ 2.5 Ghz (C/C++)
350 DEPT 21.22 % 30.85 % 18.47 % 0.03 s 1 core @ 2.5 Ghz (Python)
351 GUPNet code 21.19 % 30.29 % 18.20 % NA s 1 core @ 2.5 Ghz (Python + C/C++)
Y. Lu, X. Ma, L. Yang, T. Zhang, Y. Liu, Q. Chu, J. Yan and W. Ouyang: Geometry Uncertainty Projection Network for Monocular 3D Object Detection. arXiv preprint arXiv:2107.13774 2021.
352 HBD 20.91 % 29.87 % 18.22 % 0.12 s 1 core @ 2.5 Ghz (C/C++)
353 GPENet code 20.79 % 30.31 % 18.21 % 0.02 s GPU @ 2.5 Ghz (Python)
354 mono3d 20.75 % 31.58 % 17.66 % 0.03 s GPU @ 2.5 Ghz (Python)
355 LT-M3OD 20.74 % 29.40 % 17.83 % 0.03 s 1 core @ 2.5 Ghz (Python)
356 HomoLoss(monoflex) code 20.68 % 29.60 % 17.81 % 0.04 s 1 core @ 2.5 Ghz (Python)
J. Gu, B. Wu, L. Fan, J. Huang, S. Cao, Z. Xiang and X. Hua: Homography Loss for Monocular 3D Object Detection. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2022.
357 MonoFlex 20.67 % 30.95 % 17.72 % 0.03 s 1 core @ 2.5 Ghz (Python)
358 Anonymous 20.47 % 33.17 % 17.31 % 40 s 1 core @ 2.5 Ghz (C/C++)
359 MonoGround 20.47 % 30.07 % 17.74 % 0.03 s 1 core @ 2.5 Ghz (Python)
360 EW code 20.38 % 28.88 % 17.59 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
361 MonoDTR 20.38 % 28.59 % 17.14 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
K. Huang, T. Wu, H. Su and W. Hsu: MonoDTR: Monocular 3D Object Detection with Depth-Aware Transformer. CVPR 2022.
362 MonoEdge 20.35 % 28.80 % 17.57 % 0.05 s GPU @ 2.5 Ghz (Python)
363 SAIC_ADC_Mono3D code 20.20 % 27.09 % 18.78 % 50 s GPU @ 2.5 Ghz (Python)
364 MonoEdge-Rotate 20.16 % 31.19 % 17.35 % 0.05 s GPU @ 2.5 Ghz (Python)
365 MDSNet 20.14 % 32.81 % 15.77 % 0.07 s 1 core @ 2.5 Ghz (Python)
366 AutoShape code 20.08 % 30.66 % 15.95 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
Z. Liu, D. Zhou, F. Lu, J. Fang and L. Zhang: AutoShape: Real-Time Shape-Aware Monocular 3D Object Detection. Proceedings of the IEEE/CVF International Conference on Computer Vision 2021.
367 MonoEdge-RCNN 20.07 % 27.62 % 16.34 % 0.05 s 1 core @ 2.5 Ghz (Python)
368 M3DSSD++ code 20.03 % 32.18 % 16.47 % 0.16s 1 core @ 2.5 Ghz (C/C++)
369 MAOLoss code 19.95 % 28.29 % 16.94 % 0.05 s 1 core @ 2.5 Ghz (Python)
370 ANM 19.82 % 29.89 % 16.77 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
371 EM code 19.80 % 30.61 % 16.55 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
372 MonoFlex 19.75 % 28.23 % 16.89 % 0.03 s GPU @ 2.5 Ghz (Python)
Y. Zhang, J. Lu and J. Zhou: Objects are Different: Flexible Monocular 3D Object Detection. CVPR 2021.
373 MonoEF 19.70 % 29.03 % 17.26 % 0.03 s 1 core @ 2.5 Ghz (Python)
Y. Zhou, Y. He, H. Zhu, C. Wang, H. Li and Q. Jiang: Monocular 3D Object Detection: An Extrinsic Parameter Free Approach. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2021.
374 K3D 19.60 % 28.31 % 17.62 % 0.3 s 1 core @ 2.5 Ghz (C/C++)
375 HomoLoss(imvoxelnet) code 19.25 % 29.18 % 16.21 % 0.20 s 1 core @ 2.5 Ghz (Python)
J. Gu, B. Wu, L. Fan, J. Huang, S. Cao, Z. Xiang and X. Hua: Homogrpahy Loss for Monocular 3D Object Detection. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2022.
376 MonoAug 19.19 % 28.20 % 16.15 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
377 MK3D 19.18 % 29.11 % 15.78 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
378 DFR-Net 19.17 % 28.17 % 14.84 % 0.18 s 1080 Ti (Pytorch)
Z. Zou, X. Ye, L. Du, X. Cheng, X. Tan, L. Zhang, J. Feng, X. Xue and E. Ding: The devil is in the task: Exploiting reciprocal appearance-localization features for monocular 3d object detection . ICCV 2021.
379 SwinMono3D 19.15 % 29.65 % 14.09 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
380 DLE code 19.05 % 31.09 % 14.13 % 0.06 s NVIDIA Tesla V100
C. Liu, S. Gu, L. Gool and R. Timofte: Deep Line Encoding for Monocular 3D Object Detection and Depth Prediction. Proceedings of the British Machine Vision Conference (BMVC) 2021.
381 PCT code 19.03 % 29.65 % 15.92 % 0.045 s 1 core @ 2.5 Ghz (C/C++)
L. Wang, L. Zhang, Y. Zhu, Z. Zhang, T. He, M. Li and X. Xue: Progressive Coordinate Transforms for Monocular 3D Object Detection. NeurIPS 2021.
382 Anonymous code 18.96 % 26.54 % 16.01 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
383 CaDDN code 18.91 % 27.94 % 17.19 % 0.63 s GPU @ 2.5 Ghz (Python)
C. Reading, A. Harakeh, J. Chae and S. Waslander: Categorical Depth Distribution Network for Monocular 3D Object Detection. CVPR 2021.
384 monodle code 18.89 % 24.79 % 16.00 % 0.04 s GPU @ 2.5 Ghz (Python)
X. Ma, Y. Zhang, D. Xu, D. Zhou, S. Yi, H. Li and W. Ouyang: Delving into Localization Errors for Monocular 3D Object Detection. CVPR 2021 .
385 MonoFar 18.68 % 25.89 % 16.30 % 0.04 s 1 core @ 2.5 Ghz (Python)
386 none 18.66 % 26.19 % 15.79 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
387 Neighbor-Vote 18.65 % 27.39 % 16.54 % 0.1 s GPU @ 2.5 Ghz (Python)
X. Chu, J. Deng, Y. Li, Z. Yuan, Y. Zhang, J. Ji and Y. Zhang: Neighbor-Vote: Improving Monocular 3D Object Detection through Neighbor Distance Voting. ACM MM 2021.
388 GrooMeD-NMS code 18.27 % 26.19 % 14.05 % 0.12 s 1 core @ 2.5 Ghz (Python)
A. Kumar, G. Brazil and X. Liu: GrooMeD-NMS: Grouped Mathematically Differentiable NMS for Monocular 3D Object Detection. CVPR 2021.
389 MonoRCNN code 18.11 % 25.48 % 14.10 % 0.07 s GPU @ 2.5 Ghz (Python)
X. Shi, Q. Ye, X. Chen, C. Chen, Z. Chen and T. Kim: Geometry-based Distance Decomposition for Monocular 3D Object Detection. ICCV 2021.
390 Ground-Aware code 17.98 % 29.81 % 13.08 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
Y. Liu, Y. Yuan and M. Liu: Ground-aware Monocular 3D Object Detection for Autonomous Driving. IEEE Robotics and Automation Letters 2021.
391 MP-Mono 17.96 % 25.36 % 13.84 % 0.16 s GPU @ 2.5 Ghz (Python)
392 Aug3D-RPN 17.89 % 26.00 % 14.18 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
C. He, J. Huang, X. Hua and L. Zhang: Aug3D-RPN: Improving Monocular 3D Object Detection by Synthetic Images with Virtual Depth. 2021.
393 DDMP-3D 17.89 % 28.08 % 13.44 % 0.18 s 1 core @ 2.5 Ghz (Python)
L. Wang, L. Du, X. Ye, Y. Fu, G. Guo, X. Xue, J. Feng and L. Zhang: Depth-conditioned Dynamic Message Propagation for Monocular 3D Object Detection. CVPR 2020.
394 IAFA 17.88 % 25.88 % 15.35 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
D. Zhou, X. Song, Y. Dai, J. Yin, F. Lu, M. Liao, J. Fang and L. Zhang: IAFA: Instance-Aware Feature Aggregation for 3D Object Detection from a Single Image. Proceedings of the Asian Conference on Computer Vision 2020.
395 RefinedMPL 17.60 % 28.08 % 13.95 % 0.15 s GPU @ 2.5 Ghz (Python + C/C++)
J. Vianney, S. Aich and B. Liu: RefinedMPL: Refined Monocular PseudoLiDAR for 3D Object Detection in Autonomous Driving. arXiv preprint arXiv:1911.09712 2019.
396 Lite-FPN 17.58 % 26.67 % 14.61 % 0.01 s 1 core @ 2.5 Ghz (C/C++)
397 Kinematic3D code 17.52 % 26.69 % 13.10 % 0.12 s 1 core @ 1.5 Ghz (C/C++)
G. Brazil, G. Pons-Moll, X. Liu and B. Schiele: Kinematic 3D Object Detection in Monocular Video. ECCV 2020 .
398 MonoRUn code 17.34 % 27.94 % 15.24 % 0.07 s GPU @ 2.5 Ghz (Python + C/C++)
H. Chen, Y. Huang, W. Tian, Z. Gao and L. Xiong: MonoRUn: Monocular 3D Object Detection by Reconstruction and Uncertainty Propagation. IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2021.
399 RetinaMono 17.33 % 26.12 % 15.26 % 0.02 s 1 core @ 2.5 Ghz (Python)
400 AM3D 17.32 % 25.03 % 14.91 % 0.4 s GPU @ 2.5 Ghz (Python + C/C++)
X. Ma, Z. Wang, H. Li, P. Zhang, W. Ouyang and X. Fan: Accurate Monocular Object Detection via Color- Embedded 3D Reconstruction for Autonomous Driving. Proceedings of the IEEE international Conference on Computer Vision (ICCV) 2019.
401 YoloMono3D code 17.15 % 26.79 % 12.56 % 0.05 s GPU @ 2.5 Ghz (Python)
Y. Liu, L. Wang and L. Ming: YOLOStereo3D: A Step Back to 2D for Efficient Stereo 3D Detection. 2021 International Conference on Robotics and Automation (ICRA) 2021.
402 CMAN 17.04 % 25.89 % 12.88 % 0.15 s 1 core @ 2.5 Ghz (Python)
403 GAC3D 16.93 % 25.80 % 12.50 % 0.25 s 1 core @ 2.5 Ghz (Python)
M. Bui, D. Ngo, H. Pham and D. Nguyen: GAC3D: improving monocular 3D object detection with ground-guide model and adaptive convolution. 2021.
404 PatchNet code 16.86 % 22.97 % 14.97 % 0.4 s 1 core @ 2.5 Ghz (C/C++)
X. Ma, S. Liu, Z. Xia, H. Zhang, X. Zeng and W. Ouyang: Rethinking Pseudo-LiDAR Representation. Proceedings of the European Conference on Computer Vision (ECCV) 2020.
405 MonoAug 16.71 % 24.39 % 13.83 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
406 PGD-FCOS3D code 16.51 % 26.89 % 13.49 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
T. Wang, X. Zhu, J. Pang and D. Lin: Probabilistic and Geometric Depth: Detecting Objects in Perspective. Conference on Robot Learning (CoRL) 2021.
407 ImVoxelNet code 16.37 % 25.19 % 13.58 % 0.2 s GPU @ 2.5 Ghz (Python)
D. Rukhovich, A. Vorontsova and A. Konushin: ImVoxelNet: Image to Voxels Projection for Monocular and Multi-View General-Purpose 3D Object Detection. arXiv preprint arXiv:2106.01178 2021.
408 KM3D code 16.20 % 23.44 % 14.47 % 0.03 s 1 core @ 2.5 Ghz (Python)
P. Li: Monocular 3D Detection with Geometric Constraints Embedding and Semi-supervised Training. 2020.
409 MM 16.09 % 24.65 % 13.99 % 1 s 1 core @ 2.5 Ghz (C/C++)
410 D4LCN code 16.02 % 22.51 % 12.55 % 0.2 s GPU @ 2.5 Ghz (Python + C/C++)
M. Ding, Y. Huo, H. Yi, Z. Wang, J. Shi, Z. Lu and P. Luo: Learning Depth-Guided Convolutions for Monocular 3D Object Detection. CVPR 2020.
411 Keypoint-3D 15.54 % 23.16 % 11.83 % 14 s 1 core @ 2.5 Ghz (C/C++)
412 COF3D 15.39 % 25.36 % 11.34 % 200 s 1 core @ 2.5 Ghz (Python + C/C++)
413 MonoPair 14.83 % 19.28 % 12.89 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
Y. Chen, L. Tai, K. Sun and M. Li: MonoPair: Monocular 3D Object Detection Using Pairwise Spatial Relationships. The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2020.
414 Decoupled-3D 14.82 % 23.16 % 11.25 % 0.08 s GPU @ 2.5 Ghz (C/C++)
Y. Cai, B. Li, Z. Jiao, H. Li, X. Zeng and X. Wang: Monocular 3D Object Detection with Decoupled Structured Polygon Estimation and Height-Guided Depth Estimation. AAAI 2020.
415 QD-3DT
This is an online method (no batch processing).
 code 14.71 % 20.16 % 12.76 % 0.03 s GPU @ 2.5 Ghz (Python)
H. Hu, Y. Yang, T. Fischer, F. Yu, T. Darrell and M. Sun: Monocular Quasi-Dense 3D Object Tracking. ArXiv:2103.07351 2021.
416 SMOKE code 14.49 % 20.83 % 12.75 % 0.03 s GPU @ 2.5 Ghz (Python)
Z. Liu, Z. Wu and R. Tóth: SMOKE: Single-Stage Monocular 3D Object Detection via Keypoint Estimation. 2020.
417 RTM3D code 14.20 % 19.17 % 11.99 % 0.05 s GPU @ 1.0 Ghz (Python)
P. Li, H. Zhao, P. Liu and F. Cao: RTM3D: Real-time Monocular 3D Detection from Object Keypoints for Autonomous Driving. 2020.
418 Mono3D_PLiDAR code 13.92 % 21.27 % 11.25 % 0.1 s NVIDIA GeForce 1080 (pytorch)
X. Weng and K. Kitani: Monocular 3D Object Detection with Pseudo-LiDAR Point Cloud. arXiv:1903.09847 2019.
419 M3D-RPN code 13.67 % 21.02 % 10.23 % 0.16 s GPU @ 1.5 Ghz (Python)
G. Brazil and X. Liu: M3D-RPN: Monocular 3D Region Proposal Network for Object Detection . ICCV 2019 .
420 CSoR
This method makes use of Velodyne laser scans.
 13.07 % 18.67 % 10.34 % 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.
421 MonoPSR code 12.58 % 18.33 % 9.91 % 0.2 s GPU @ 3.5 Ghz (Python)
J. Ku*, A. Pon* and S. Waslander: Monocular 3D Object Detection Leveraging Accurate Proposals and Shape Reconstruction. CVPR 2019.
422 MonoCInIS 11.64 % 22.28 % 9.95 % 0,13 s GPU @ 2.5 Ghz (C/C++)
J. Heylen, M. De Wolf, B. Dawagne, M. Proesmans, L. Van Gool, W. Abbeloos, H. Abdelkawy and D. Reino: MonoCInIS: Camera Independent Monocular 3D Object Detection using Instance Segmentation. Proceedings of the IEEE/CVF International Conference on Computer Vision 2021.
423 SS3D 11.52 % 16.33 % 9.93 % 48 ms Tesla V100 (Python)
E. Jörgensen, C. Zach and F. Kahl: Monocular 3D Object Detection and Box Fitting Trained End-to-End Using Intersection-over-Union Loss. CoRR 2019.
424 MonoGRNet code 11.17 % 18.19 % 8.73 % 0.04s NVIDIA P40
Z. Qin, J. Wang and Y. Lu: MonoGRNet: A Geometric Reasoning Network for 3D Object Localization. The Thirty-Third AAAI Conference on Artificial Intelligence (AAAI-19) 2019.
425 MonoFENet 11.03 % 17.03 % 9.05 % 0.15 s 1 core @ 3.5 Ghz (Python)
W. Bao, B. Xu and Z. Chen: MonoFENet: Monocular 3D Object Detection with Feature Enhancement Networks. IEEE Transactions on Image Processing 2019.
426 MonoCInIS 10.96 % 20.42 % 9.23 % 0,14 s GPU @ 2.5 Ghz (Python)
J. Heylen, M. De Wolf, B. Dawagne, M. Proesmans, L. Van Gool, W. Abbeloos, H. Abdelkawy and D. Reino: MonoCInIS: Camera Independent Monocular 3D Object Detection using Instance Segmentation. Proceedings of the IEEE/CVF International Conference on Computer Vision 2021.
427 A3DODWTDA (image) code 8.66 % 10.37 % 7.06 % 0.8 s GPU @ 3.0 Ghz (Python)
F. Gustafsson and E. Linder-Norén: Automotive 3D Object Detection Without Target Domain Annotations. 2018.
428 TLNet (Stereo)
This method uses stereo information.
 code 7.69 % 13.71 % 6.73 % 0.1 s 1 core @ 2.5 Ghz (Python)
Z. Qin, J. Wang and Y. Lu: Triangulation Learning Network: from Monocular to Stereo 3D Object Detection. IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2019.
429 Shift R-CNN (mono) code 6.82 % 11.84 % 5.27 % 0.25 s GPU @ 1.5 Ghz (Python)
A. Naiden, V. Paunescu, G. Kim, B. Jeon and M. Leordeanu: Shift R-CNN: Deep Monocular 3D Object Detection With Closed-form Geometric Constraints. ICIP 2019.
430 SparVox3D 6.39 % 10.20 % 5.06 % 0.05 s GPU @ 2.0 Ghz (Python)
E. Balatkan and F. Kıraç: Improving Regression Performance on Monocular 3D Object Detection Using Bin-Mixing and Sparse Voxel Data. 2021 6th International Conference on Computer Science and Engineering (UBMK) 2021.
431 GS3D 6.08 % 8.41 % 4.94 % 2 s 1 core @ 2.5 Ghz (C/C++)
B. Li, W. Ouyang, L. Sheng, X. Zeng and X. Wang: GS3D: An Efficient 3D Object Detection Framework for Autonomous Driving. IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2019.
432 MVRA + I-FRCNN+ 5.84 % 9.05 % 4.50 % 0.18 s GPU @ 2.5 Ghz (Python)
H. Choi, H. Kang and Y. Hyun: Multi-View Reprojection Architecture for Orientation Estimation. The IEEE International Conference on Computer Vision (ICCV) Workshops 2019.
433 WeakM3D code 5.66 % 11.82 % 4.08 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
L. Peng, S. Yan, B. Wu, Z. Yang, X. He and D. Cai: WeakM3D: Towards Weakly Supervised Monocular 3D Object Detection. ICLR 2022.
434 ROI-10D 4.91 % 9.78 % 3.74 % 0.2 s GPU @ 3.5 Ghz (Python)
F. Manhardt, W. Kehl and A. Gaidon: ROI-10D: Monocular Lifting of 2D Detection to 6D Pose and Metric Shape. Computer Vision and Pattern Recognition (CVPR) 2019.
435 CDTrack3D code 4.61 % 7.02 % 3.73 % 0.0106 s NVIDIA RTX 3090 GPU, i9 10850k CPU
436 3D-GCK 4.57 % 5.79 % 3.64 % 24 ms Tesla V100
N. Gählert, J. Wan, N. Jourdan, J. Finkbeiner, U. Franke and J. Denzler: Single-Shot 3D Detection of Vehicles from Monocular RGB Images via Geometrically Constrained Keypoints in Real-Time. 2020 IEEE Intelligent Vehicles Symposium (IV) 2020.
437 FQNet 3.23 % 5.40 % 2.46 % 0.5 s 1 core @ 2.5 Ghz (Python)
L. Liu, J. Lu, C. Xu, Q. Tian and J. Zhou: Deep Fitting Degree Scoring Network for Monocular 3D Object Detection. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2019.
438 3D-SSMFCNN code 2.63 % 3.20 % 2.40 % 0.1 s GPU @ 1.5 Ghz (C/C++)
L. Novak: Vehicle Detection and Pose Estimation for Autonomous Driving. 2017.
439 VeloFCN
This method makes use of Velodyne laser scans.
 0.14 % 0.02 % 0.21 % 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 .
440 MonoDET code 0.14 % 0.25 % 0.10 % 0.04 s 1 core @ 2.5 Ghz (Python)
441 test code 0.09 % 0.04 % 0.11 % 50 s 1 core @ 2.5 Ghz (Python)
442 multi-task CNN 0.00 % 0.00 % 0.00 % 25.1 ms GPU @ 2.0 Ghz (Python)
M. Oeljeklaus, F. Hoffmann and T. Bertram: A Fast Multi-Task CNN for Spatial Understanding of Traffic Scenes. IEEE Intelligent Transportation Systems Conference 2018.
443 mBoW
This method makes use of Velodyne laser scans.
 0.00 % 0.00 % 0.00 % 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.
Table as LaTeX | Only published Methods

Pedestrian


Method Setting Code Moderate Easy Hard Runtime Environment
1 PiFeNet 53.92 % 63.25 % 50.53 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
D. Le, H. Shi, H. Rezatofighi and J. Cai: PiFeNet: Pillar-Feature Network for Real-Time 3D Pedestrian Detection from Point Cloud. 2021.
2 CasA++ 53.84 % 60.14 % 51.35 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
3 TED 53.48 % 60.13 % 50.89 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
4 DCAN-Second code 53.18 % 60.92 % 50.56 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
5 EQ-PVRCNN code 52.81 % 61.73 % 49.87 % 0.2 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Yang, L. Jiang, Y. Sun, B. Schiele and J. Jia: A Unified Query-based Paradigm for Point Cloud Understanding. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2022.
6 PV-RCNN++ code 52.43 % 59.73 % 48.73 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
7 VPFNet code 52.41 % 60.07 % 50.28 % 0.2 s 1 core @ 2.5 Ghz (C/C++)
C. Wang, H. Chen and L. Fu: VPFNet: Voxel-Pixel Fusion Network for Multi-class 3D Object Detection. 2021.
8 Frustum-PointPillars code 52.23 % 60.98 % 48.30 % 0.06 s 4 cores @ 3.0 Ghz (Python)
A. Paigwar, D. Sierra-Gonzalez, \. Erkent and C. Laugier: Frustum-PointPillars: A Multi-Stage Approach for 3D Object Detection using RGB Camera and LiDAR. International Conference on Computer Vision, ICCV, Workshop on Autonomous Vehicle Vision 2021.
9 CAD 52.20 % 60.23 % 49.54 % 0.1 s GPU @ 2.5 Ghz (Python + C/C++)
10 TANet code 51.38 % 60.85 % 47.54 % 0.035s GPU @ 2.5 Ghz (Python + C/C++)
Z. Liu, X. Zhao, T. Huang, R. Hu, Y. Zhou and X. Bai: TANet: Robust 3D Object Detection from Point Clouds with Triple Attention. AAAI 2020.
11 CasA 51.37 % 57.95 % 49.08 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
12 ISE-RCNN 51.06 % 55.64 % 47.76 % 0.09 s 1 core @ 2.5 Ghz (Python + C/C++)
13 MMLab PV-RCNN
This method makes use of Velodyne laser scans.
 code 50.57 % 59.86 % 46.74 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
S. Shi, C. Guo, L. Jiang, Z. Wang, J. Shi, X. Wang and H. Li: PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detection. CVPR 2020.
14 HotSpotNet 50.53 % 57.39 % 46.65 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
Q. Chen, L. Sun, Z. Wang, K. Jia and A. Yuille: object as hotspots. Proceedings of the European Conference on Computer Vision (ECCV) 2020.
15 VMVS
This method makes use of Velodyne laser scans.
 50.34 % 60.34 % 46.45 % 0.25 s GPU @ 2.5 Ghz (Python)
J. Ku, A. Pon, S. Walsh and S. Waslander: Improving 3D object detection for pedestrians with virtual multi-view synthesis orientation estimation. IROS 2019.
16 AVOD-FPN
This method makes use of Velodyne laser scans.
 code 50.32 % 58.49 % 46.98 % 0.1 s Titan X (Pascal)
J. Ku, M. Mozifian, J. Lee, A. Harakeh and S. Waslander: Joint 3D Proposal Generation and Object Detection from View Aggregation. IROS 2018.
17 SPT 50.22 % 56.54 % 46.72 % 0.1 s GPU @ 2.5 Ghz (Python)
18 variance_point 50.03 % 57.72 % 46.27 % 0.05 s 1 core @ 2.5 Ghz (Python)
19 3DSSD code 49.94 % 60.54 % 45.73 % 0.04 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Yang, Y. Sun, S. Liu and J. Jia: 3DSSD: Point-based 3D Single Stage Object Detector. CVPR 2020.
20 PointPainting
This method makes use of Velodyne laser scans.
 49.93 % 58.70 % 46.29 % 0.4 s GPU @ 2.5 Ghz (Python + C/C++)
S. Vora, A. Lang, B. Helou and O. Beijbom: PointPainting: Sequential Fusion for 3D Object Detection. CVPR 2020.
21 SemanticVoxels 49.93 % 58.91 % 47.31 % 0.04 s GPU @ 2.5 Ghz (Python + C/C++)
J. Fei, W. Chen, P. Heidenreich, S. Wirges and C. Stiller: SemanticVoxels: Sequential Fusion for 3D Pedestrian Detection using LiDAR Point Cloud and Semantic Segmentation. MFI 2020.
22 MMLab-PartA^2
This method makes use of Velodyne laser scans.
 code 49.81 % 59.04 % 45.92 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
S. Shi, Z. Wang, J. Shi, X. Wang and H. Li: From Points to Parts: 3D Object Detection from Point Cloud with Part-aware and Part-aggregation Network. IEEE Transactions on Pattern Analysis and Machine Intelligence 2020.
23 TBD 49.59 % 58.17 % 47.08 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
24 F-PointNet
This method makes use of Velodyne laser scans.
 code 49.57 % 57.13 % 45.48 % 0.17 s GPU @ 3.0 Ghz (Python)
C. Qi, W. Liu, C. Wu, H. Su and L. Guibas: Frustum PointNets for 3D Object Detection from RGB-D Data. arXiv preprint arXiv:1711.08488 2017.
25 TBD 49.56 % 58.10 % 47.05 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
26 CFF-tv 49.29 % 57.83 % 46.70 % 1 s 1 core @ 2.5 Ghz (C/C++)
27 AutoAlign 49.27 % 59.28 % 46.14 % 0.1 s 1 core @ 2.5 Ghz (Python)
28 VPN 49.19 % 57.98 % 45.26 % 0.06 s 1 core @ 2.5 Ghz (Python + C/C++)
29 F-ConvNet
This method makes use of Velodyne laser scans.
 code 48.96 % 57.04 % 44.33 % 0.47 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Wang and K. Jia: Frustum ConvNet: Sliding Frustums to Aggregate Local Point-Wise Features for Amodal 3D Object Detection. IROS 2019.
30 HVNet 48.86 % 54.84 % 46.33 % 0.03 s GPU @ 2.0 Ghz (Python)
M. Ye, S. Xu and T. Cao: HVNet: Hybrid Voxel Network for LiDAR Based 3D Object Detection. CVPR 2020.
31 CAT-Det 48.78 % 57.13 % 45.56 % 0.3 s GPU @ 2.5 Ghz (Python + C/C++)
Y. Zhang, J. Chen and D. Huang: CAT-Det: Contrastively Augmented Transformer for Multi-modal 3D Object Detection. CVPR 2022.
32 PE-RCVN 48.72 % 54.09 % 46.71 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
33 STD code 48.72 % 60.02 % 44.55 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Yang, Y. Sun, S. Liu, X. Shen and J. Jia: STD: Sparse-to-Dense 3D Object Detector for Point Cloud. ICCV 2019.
34 Reprod-Two-Branch 48.71 % 57.25 % 45.75 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
35 KeyFuse2B 48.64 % 56.16 % 46.20 % 0.01 s 1 core @ 2.5 Ghz (C/C++)
36 PointPillars
This method makes use of Velodyne laser scans.
 code 48.64 % 57.60 % 45.78 % 16 ms 1080ti GPU and Intel i7 CPU
A. Lang, S. Vora, H. Caesar, L. Zhou, J. Yang and O. Beijbom: PointPillars: Fast Encoders for Object Detection from Point Clouds. CVPR 2019.
37 USVLab BSAODet 48.61 % 55.76 % 46.08 % 0.04 s GPU @ 2.5 Ghz (Python + C/C++)
38 FV2P v2 48.58 % 54.90 % 45.11 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
39 FS-Net
This method makes use of Velodyne laser scans.
 48.50 % 54.91 % 46.29 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
40 EPNet++ 48.47 % 56.24 % 45.73 % 0.1 s GPU @ 2.5 Ghz (Python)
Z. Liu, H. tengteng, B. Li, X. Chen, X. Wang and X. Bai: EPNet++: Cascade Bi-directional Fusion for Multi-Modal 3D Object Detection. arXiv preprint arXiv:2112.11088 2021.
41 MGAF-3DSSD code 48.46 % 56.09 % 44.90 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
J. Li, H. Dai, L. Shao and Y. Ding: Anchor-free 3D Single Stage Detector with Mask-Guided Attention for Point Cloud. MM '21: The 29th ACM International Conference on Multimedia (ACM MM) 2021.
42 CFF-ep25 48.31 % 56.34 % 45.58 % 1 s 1 core @ 2.5 Ghz (C/C++)
43 Fast-CLOCs 48.27 % 57.19 % 44.55 % 0.1 s GPU @ 2.5 Ghz (Python)
S. Pang, D. Morris and H. Radha: Fast-CLOCs: Fast Camera-LiDAR Object Candidates Fusion for 3D Object Detection. Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) 2022.
44 FromVoxelToPoint code 48.15 % 56.54 % 45.63 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
J. Li, H. Dai, L. Shao and Y. Ding: From Voxel to Point: IoU-guided 3D Object Detection for Point Cloud with Voxel-to- Point Decoder. MM '21: The 29th ACM International Conference on Multimedia (ACM MM) 2021.
45 cff-tv-v2-ep25 48.13 % 56.48 % 45.66 % 1 s 1 core @ 2.5 Ghz (C/C++)
46 USVLab BSAODet (S) 48.10 % 54.96 % 45.65 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
47 ISE-RCNN-PV 47.85 % 55.63 % 45.80 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
48 tbd 47.84 % 57.69 % 43.99 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
49 HMFI code 47.77 % 55.61 % 45.17 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
50 TBD 47.77 % 55.61 % 45.17 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
51 CFF-tv-v2 47.59 % 55.46 % 45.09 % 1 s 1 core @ 2.5 Ghz (C/C++)
52 Anonymous 47.47 % 52.94 % 45.41 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
53 CF-ctdep-tv-ta 47.46 % 54.36 % 45.07 % 1 s 1 core @ 2.5 Ghz (C/C++)
54 P2V-RCNN 47.36 % 54.15 % 45.10 % 0.1 s 2 cores @ 2.5 Ghz (Python)
J. Li, S. Luo, Z. Zhu, H. Dai, A. Krylov, Y. Ding and L. Shao: P2V-RCNN: Point to Voxel Feature Learning for 3D Object Detection from Point Clouds. IEEE Access 2021.
55 SGNet 47.29 % 53.84 % 44.10 % 0.09 s GPU @ 2.5 Ghz (Python)
56 CF-base-tv 47.28 % 54.77 % 44.81 % 1 s 1 core @ 2.5 Ghz (C/C++)
57 Self-Calib Conv 47.17 % 54.20 % 44.84 % 0.01 s 1 core @ 2.5 Ghz (C/C++)
58 TCDVF 47.11 % 55.26 % 44.53 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
59 Point-GNN
This method makes use of Velodyne laser scans.
 code 47.07 % 55.36 % 44.61 % 0.6 s GPU @ 2.5 Ghz (Python)
W. Shi and R. Rajkumar: Point-GNN: Graph Neural Network for 3D Object Detection in a Point Cloud. CVPR 2020.
60 MVMM code 46.84 % 53.75 % 44.87 % 0.04 s GPU @ 2.5 Ghz (Python + C/C++)
61 SCNet
This method makes use of Velodyne laser scans.
 46.73 % 56.87 % 42.74 % 0.04 s GPU @ 3.0 Ghz (Python)
Z. Wang, H. Fu, L. Wang, L. Xiao and B. Dai: SCNet: Subdivision Coding Network for Object Detection Based on 3D Point Cloud. IEEE Access 2019.
62 cp-tv-kp 46.71 % 53.73 % 44.45 % 1 s 1 core @ 2.5 Ghz (C/C++)
63 DGT-Det3D code 46.59 % 54.25 % 44.15 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
64 PSA-Det3D 46.36 % 53.26 % 43.73 % 0.1 s GPU @ 2.5 Ghz (Python)
65 CF-ctdep-tv 46.36 % 53.50 % 44.01 % 1 s 1 core @ 2.5 Ghz (C/C++)
66 MSADet 46.27 % 55.91 % 43.83 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
67 3SNet 46.25 % 52.22 % 42.89 % 0.07 s GPU @ 2.5 Ghz (Python)
68 DGT-Det3D 46.22 % 53.98 % 43.85 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
69 MMLab-PointRCNN
This method makes use of Velodyne laser scans.
 code 46.13 % 54.77 % 42.84 % 0.1 s GPU @ 2.5 Ghz (Python + C/C++)
S. Shi, X. Wang and H. Li: Pointrcnn: 3d object proposal generation and detection from point cloud. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2019.
70 ARPNET 45.92 % 55.48 % 42.54 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
Y. Ye, C. Zhang and X. Hao: ARPNET: attention region proposal network for 3D object detection. Science China Information Sciences 2019.
71 CenterFuse 45.84 % 55.20 % 43.46 % 0.059 sec/frame 2 x V100
72 DSA-PV-RCNN
This method makes use of Velodyne laser scans.
 code 45.82 % 52.03 % 43.81 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
P. Bhattacharyya, C. Huang and K. Czarnecki: SA-Det3D: Self-Attention Based Context-Aware 3D Object Detection. 2021.
73 cp-tv 45.75 % 52.90 % 43.49 % 1 s 1 core @ 2.5 Ghz (C/C++)
74 SVGA-Net 45.68 % 53.09 % 43.30 % 0.03s 1 core @ 2.5 Ghz (Python + C/C++)
Q. He, Z. Wang, H. Zeng, Y. Zeng and Y. Liu: SVGA-Net: Sparse Voxel-Graph Attention Network for 3D Object Detection from Point Clouds. AAAI 2022.
75 SARFE 45.60 % 51.45 % 43.49 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
76 TBD 45.57 % 52.08 % 42.35 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
77 epBRM
This method makes use of Velodyne laser scans.
 code 45.49 % 52.48 % 42.75 % 0.10 s 1 core @ 2.5 Ghz (C/C++)
K. Shin: Improving a Quality of 3D Object Detection by Spatial Transformation Mechanism. arXiv preprint arXiv:1910.04853 2019.
78 TBD code 45.46 % 52.72 % 42.53 % 0.01 s 1 core @ 2.5 Ghz (C/C++)
79 PDV code 45.45 % 51.95 % 43.33 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
J. Hu, T. Kuai and S. Waslander: Point Density-Aware Voxels for LiDAR 3D Object Detection. CVPR 2022.
80 MLOD
This method makes use of Velodyne laser scans.
 code 45.40 % 55.09 % 41.42 % 0.12 s GPU @ 1.5 Ghz (Python)
J. Deng and K. Czarnecki: MLOD: A multi-view 3D object detection based on robust feature fusion method. arXiv preprint arXiv:1909.04163 2019.
81 cp-tv-kp-io-sc 45.30 % 53.84 % 42.12 % 1 s 1 core @ 2.5 Ghz (C/C++)
82 IA-SSD (single) code 45.07 % 52.73 % 42.75 % 0.013 s 1 core @ 2.5 Ghz (C/C++)
Y. Zhang, Q. Hu, G. Xu, Y. Ma, J. Wan and Y. Guo: Not All Points Are Equal: Learning Highly Efficient Point-based Detectors for 3D LiDAR Point Clouds. CVPR 2022.
83 TBD 44.99 % 50.41 % 42.85 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
84 FusionDetv1 44.85 % 52.42 % 42.56 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
85 SRDL 44.84 % 52.42 % 42.56 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
ERROR: Wrong syntax in BIBTEX file.
86 M3DeTR code 44.78 % 50.63 % 42.57 % n/a s GPU @ 1.0 Ghz (Python)
T. Guan, J. Wang, S. Lan, R. Chandra, Z. Wu, L. Davis and D. Manocha: M3DeTR: Multi-representation, Multi- scale, Mutual-relation 3D Object Detection with Transformers. 2021.
87 WGVRF 44.75 % 50.80 % 42.78 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
88 Semantical PVRCNN 44.75 % 49.40 % 41.94 % 0.07 s 1 core @ 2.5 Ghz (C/C++)
89 AFTD 44.74 % 53.94 % 42.36 % 1 s 1 core @ 2.5 Ghz (Python + C/C++)
90 AGS-SSD[la] 44.65 % 51.26 % 41.57 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
91 FusionDetv2-v5 44.64 % 51.44 % 42.32 % 0.05 s 1 core @ 2.5 Ghz (Java + C/C++)
92 Dune-DCF-e11 44.58 % 52.44 % 41.75 % 1 s 1 core @ 2.5 Ghz (C/C++)
93 ATT_SSD 44.57 % 51.26 % 42.33 % 0.01 s 1 core @ 2.5 Ghz (Python)
94 CF-cd-io-tv 44.54 % 53.64 % 41.21 % 1 s 1 core @ 2.5 Ghz (C/C++)
95 Dune-DCF-e09 44.50 % 52.64 % 41.86 % 1 s 1 core @ 2.5 Ghz (C/C++)
96 FusionDetv2-v4 44.47 % 50.88 % 42.18 % 0.06 s 1 core @ 2.5 Ghz (Python + C/C++)
97 IKT3D
This method makes use of Velodyne laser scans.
 44.45 % 49.50 % 42.58 % 0.05 s 1 core @ 2.5 Ghz (Python)
98 SIF 44.28 % 52.05 % 42.03 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
P. An: SIF. Submitted to CVIU 2021.
99 LazyTorch-CP-Infer-O 44.27 % 51.92 % 41.99 % 1 s 1 core @ 2.5 Ghz (C/C++)
100 KeyPoint-IoUHead 44.27 % 53.12 % 41.83 % 0.01 s 1 core @ 2.5 Ghz (C/C++)
101 LazyTorch-CP-Small-P 44.25 % 51.84 % 41.97 % 1 s 1 core @ 2.5 Ghz (C/C++)
102 DDet 44.24 % 50.01 % 42.38 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
103 IoU-2B 44.19 % 55.31 % 40.33 % 0.01 s 1 core @ 2.5 Ghz (C/C++)
104 ST-RCNN
This method makes use of Velodyne laser scans.
 44.14 % 49.78 % 41.95 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
105 DVFENet 44.12 % 50.98 % 41.62 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
Y. He, G. Xia, Y. Luo, L. Su, Z. Zhang, W. Li and P. Wang: DVFENet: Dual-branch Voxel Feature Extraction Network for 3D Object Detection. Neurocomputing 2021.
106 VCRCNN 44.09 % 48.82 % 42.19 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
107 CenterPoint (pcdet) 44.08 % 51.76 % 41.80 % 0.051 sec/frame 2 x V100
108 CrazyTensor-CP 44.06 % 51.25 % 41.50 % 1 s 1 core @ 2.5 Ghz (Python)
109 cff-tv-t 44.00 % 54.42 % 41.46 % 1 s 1 core @ 2.5 Ghz (C/C++)
110 DSASNet 43.98 % 50.55 % 40.63 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
111 CF-base-train 43.90 % 51.40 % 41.24 % 1 s 1 core @ 2.5 Ghz (C/C++)
112 City-CF-fixed 43.86 % 51.92 % 41.33 % 1 s 1 core @ 2.5 Ghz (C/C++)
113 Faraway-Frustum
This method makes use of Velodyne laser scans.
 code 43.85 % 52.15 % 41.68 % 0.1 s GPU @ 2.5 Ghz (Python)
H. Zhang, D. Yang, E. Yurtsever, K. Redmill and U. Ozguner: Faraway-frustum: Dealing with lidar sparsity for 3D object detection using fusion. 2021 IEEE International Intelligent Transportation Systems Conference (ITSC) 2021.
114 Dune-DCF-e15 43.63 % 51.18 % 41.11 % 1 s 1 core @ 2.5 Ghz (C/C++)
115 S-AT GCN 43.43 % 50.63 % 41.58 % 0.02 s GPU @ 2.0 Ghz (Python)
L. Wang, C. Wang, X. Zhang, T. Lan and J. Li: S-AT GCN: Spatial-Attention Graph Convolution Network based Feature Enhancement for 3D Object Detection. CoRR 2021.
116 CF-ctdep-train 43.20 % 50.14 % 40.69 % 1 s 1 core @ 2.5 Ghz (C/C++)
117 FPV-SSD 43.19 % 50.37 % 40.95 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
118 City-CF 42.95 % 49.91 % 40.61 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
119 BirdNet+
This method makes use of Velodyne laser scans.
 code 42.87 % 48.90 % 40.59 % 0.11 s Titan Xp (PyTorch)
A. Barrera, J. Beltrán, C. Guindel, J. Iglesias and F. García: BirdNet+: Two-Stage 3D Object Detection in LiDAR through a Sparsity-Invariant Bird’s Eye View. IEEE Access 2021.
120 CZY 42.80 % 49.42 % 40.83 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
121 TBD 42.76 % 50.17 % 39.75 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
122 IA-SSD (multi) code 42.61 % 51.76 % 40.51 % 0.014 s 1 core @ 2.5 Ghz (C/C++)
Y. Zhang, Q. Hu, G. Xu, Y. Ma, J. Wan and Y. Guo: Not All Points Are Equal: Learning Highly Efficient Point-based Detectors for 3D LiDAR Point Clouds. CVPR 2022.
123 HS3D code 42.60 % 51.58 % 39.27 % 0.12 s 1 core @ 2.5 Ghz (Python + C/C++)
124 NV-RCNN 42.58 % 49.00 % 40.39 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
125 FusionDetv2-baseline 42.53 % 47.08 % 40.71 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
126 XView 42.42 % 47.24 % 39.96 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
L. Xie, G. Xu, D. Cai and X. He: X-view: Non-egocentric Multi-View 3D Object Detector. 2021.
127 PVTr 42.26 % 48.79 % 40.27 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
128 T_PVRCNN_V2 42.21 % 50.58 % 39.81 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
129 T_PVRCNN 41.87 % 49.87 % 39.44 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
130 PFF3D
This method makes use of Velodyne laser scans.
 code 40.94 % 48.74 % 38.54 % 0.05 s GPU @ 3.0 Ghz (Python + C/C++)
L. Wen and K. Jo: Fast and Accurate 3D Object Detection for Lidar-Camera-Based Autonomous Vehicles Using One Shared Voxel-Based Backbone. IEEE Access 2021.
131 CrazyTensor-CF 40.78 % 48.79 % 38.16 % 1 s 1 core @ 2.5 Ghz (C/C++)
132 NV2P-RCNN 40.71 % 46.83 % 38.86 % 0.1 s GPU @ 2.5 Ghz (Python)
133 TBD_BD code 40.41 % 48.27 % 38.45 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
134 ZMMPP 39.11 % 46.50 % 37.04 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
135 DSGN++
This method uses stereo information.
 code 38.92 % 50.26 % 35.12 % 0.2 s GeForce RTX 2080Ti
Y. Chen, S. Huang, S. Liu, B. Yu and J. Jia: DSGN++: Exploiting Visual-Spatial Relation for Stereo-based 3D Detectors. arXiv preprint arXiv:2204.03039 2022.
136 AB3DMOT
This method makes use of Velodyne laser scans.
This is an online method (no batch processing).
 code 38.79 % 47.51 % 35.85 % 0.0047s 1 core @ 2.5 Ghz (python)
X. Weng and K. Kitani: A Baseline for 3D Multi-Object Tracking. arXiv:1907.03961 2019.
137 BirdNet+ (legacy)
This method makes use of Velodyne laser scans.
 code 38.28 % 45.53 % 35.37 % 0.1 s Titan Xp (PyTorch)
A. Barrera, C. Guindel, J. Beltrán and F. García: BirdNet+: End-to-End 3D Object Detection in LiDAR Bird’s Eye View. 2020 IEEE 23rd International Conference on Intelligent Transportation Systems (ITSC) 2020.
138 PP-PCdet code 38.21 % 45.14 % 36.04 % 0.01 s 1 core @ 2.5 Ghz (Python)
139 CSW3D
This method makes use of Velodyne laser scans.
 37.96 % 49.27 % 33.83 % 0.03 s 4 cores @ 2.5 Ghz (C/C++)
J. Hu, T. Wu, H. Fu, Z. Wang and K. Ding: Cascaded Sliding Window Based Real-Time 3D Region Proposal for Pedestrian Detection. ROBIO 2019.
140 KPP3D code 37.82 % 45.25 % 35.36 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
141 StereoDistill 37.75 % 50.79 % 34.28 % 0.4 s 1 core @ 2.5 Ghz (Python)
142 Contrastive PP code 37.68 % 44.10 % 35.84 % 0.01 s 1 core @ 2.5 Ghz (Python)
143 CZY_3917 37.26 % 44.66 % 34.56 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
144 GT3D 36.37 % 46.02 % 32.62 % 0.1 s 1 core @ 2.5 Ghz (Python)
145 DMF
This method uses stereo information.
 34.92 % 42.08 % 32.69 % 0.2 s 1 core @ 2.5 Ghz (Python + C/C++)
X. J. Chen and W. Xu: Disparity-Based Multiscale Fusion Network for Transportation Detection. IEEE Transactions on Intelligent Transportation Systems 2022.
146 SparsePool code 34.15 % 43.33 % 31.78 % 0.13 s 8 cores @ 2.5 Ghz (Python)
Z. Wang, W. Zhan and M. Tomizuka: Fusing bird view lidar point cloud and front view camera image for deep object detection. arXiv preprint arXiv:1711.06703 2017.
147 MMLAB LIGA-Stereo
This method uses stereo information.
 code 34.13 % 44.71 % 30.42 % 0.4 s 1 core @ 2.5 Ghz (Python + C/C++)
X. Guo, S. Shi, X. Wang and H. Li: LIGA-Stereo: Learning LiDAR Geometry Aware Representations for Stereo-based 3D Detector. Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) 2021.
148 AVOD
This method makes use of Velodyne laser scans.
 code 33.57 % 42.58 % 30.14 % 0.08 s Titan X (pascal)
J. Ku, M. Mozifian, J. Lee, A. Harakeh and S. Waslander: Joint 3D Proposal Generation and Object Detection from View Aggregation. IROS 2018.
149 SparsePool code 33.22 % 41.55 % 29.66 % 0.13 s 8 cores @ 2.5 Ghz (Python)
Z. Wang, W. Zhan and M. Tomizuka: Fusing bird view lidar point cloud and front view camera image for deep object detection. arXiv preprint arXiv:1711.06703 2017.
150 PS++ code 32.38 % 43.37 % 28.66 % PS++ s 1 core @ 2.5 Ghz (C/C++)
151 CZY 32.05 % 39.50 % 29.90 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
152 PS code 31.13 % 41.55 % 27.50 % PS s 1 core @ 2.5 Ghz (C/C++)
153 CG-Stereo
This method uses stereo information.
 29.56 % 39.24 % 25.87 % 0.57 s GeForce RTX 2080 Ti
C. Li, J. Ku and S. Waslander: Confidence Guided Stereo 3D Object Detection with Split Depth Estimation. IROS 2020.
154 PointRGBNet 29.32 % 38.07 % 26.94 % 0.08 s 4 cores @ 2.5 Ghz (Python + C/C++)
P. Xie Desheng: Real-time Detection of 3D Objects Based on Multi-Sensor Information Fusion. Automotive Engineering 2022.
155 Disp R-CNN
This method uses stereo information.
 code 29.12 % 42.72 % 25.09 % 0.387 s GPU @ 2.5 Ghz (Python + C/C++)
J. Sun, L. Chen, Y. Xie, S. Zhang, Q. Jiang, X. Zhou and H. Bao: Disp R-CNN: Stereo 3D Object Detection via Shape Prior Guided Instance Disparity Estimation. CVPR 2020.
156 Disp R-CNN (velo)
This method uses stereo information.
 code 28.34 % 40.21 % 24.46 % 0.387 s GPU @ 2.5 Ghz (Python + C/C++)
J. Sun, L. Chen, Y. Xie, S. Zhang, Q. Jiang, X. Zhou and H. Bao: Disp R-CNN: Stereo 3D Object Detection via Shape Prior Guided Instance Disparity Estimation. CVPR 2020.
157 BirdNet
This method makes use of Velodyne laser scans.
 23.06 % 28.20 % 21.65 % 0.11 s Titan Xp (Caffe)
J. Beltrán, C. Guindel, F. Moreno, D. Cruzado, F. García and A. Escalera: BirdNet: A 3D Object Detection Framework from LiDAR Information. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
158 AEC3D 22.40 % 28.59 % 20.67 % 18 ms GPU @ 2.5 Ghz (Python)
159 OC Stereo
This method uses stereo information.
 code 20.80 % 29.79 % 18.62 % 0.35 s 1 core @ 2.5 Ghz (Python + C/C++)
A. Pon, J. Ku, C. Li and S. Waslander: Object-Centric Stereo Matching for 3D Object Detection. ICRA 2020.
160 YOLOStereo3D
This method uses stereo information.
 code 20.76 % 31.01 % 18.41 % 0.1 s GPU 1080Ti
Y. Liu, L. Wang and M. Liu: YOLOStereo3D: A Step Back to 2D for Efficient Stereo 3D Detection. 2021 International Conference on Robotics and Automation (ICRA) 2021.
161 DSGN
This method uses stereo information.
 code 20.75 % 26.61 % 18.86 % 0.67 s NVIDIA Tesla V100
Y. Chen, S. Liu, X. Shen and J. Jia: DSGN: Deep Stereo Geometry Network for 3D Object Detection. CVPR 2020.
162 Complexer-YOLO
This method makes use of Velodyne laser scans.
 18.26 % 21.42 % 17.06 % 0.06 s GPU @ 3.5 Ghz (C/C++)
M. Simon, K. Amende, A. Kraus, J. Honer, T. Samann, H. Kaulbersch, S. Milz and H. Michael Gross: Complexer-YOLO: Real-Time 3D Object Detection and Tracking on Semantic Point Clouds. The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops 2019.
163 TopNet-Retina
This method makes use of Velodyne laser scans.
 14.57 % 18.04 % 12.48 % 52ms GeForce 1080Ti (tensorflow-gpu, v1.12)
S. Wirges, T. Fischer, C. Stiller and J. Frias: Object Detection and Classification in Occupancy Grid Maps Using Deep Convolutional Networks. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
164 RT3D-GMP
This method uses stereo information.
 14.22 % 19.92 % 12.83 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
H. Königshof and C. Stiller: Learning-Based Shape Estimation with Grid Map Patches for Realtime 3D Object Detection for Automated Driving. 2020 IEEE 23rd International Conference on Intelligent Transportation Systems (ITSC) 2020.
165 TopNet-HighRes
This method makes use of Velodyne laser scans.
 13.50 % 19.43 % 11.93 % 101ms NVIDIA GeForce 1080 Ti (tensorflow-gpu)
S. Wirges, T. Fischer, C. Stiller and J. Frias: Object Detection and Classification in Occupancy Grid Maps Using Deep Convolutional Networks. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
166 Anonymous 13.47 % 20.42 % 11.64 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
167 ESGN
This method uses stereo information.
 13.03 % 17.94 % 11.54 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
168 DD3D code 12.51 % 18.58 % 10.65 % n/a s 1 core @ 2.5 Ghz (C/C++)
D. Park, R. Ambrus, V. Guizilini, J. Li and A. Gaidon: Is Pseudo-Lidar needed for Monocular 3D Object detection?. IEEE/CVF International Conference on Computer Vision (ICCV) .
169 DEPT 12.29 % 18.05 % 10.50 % 0.03 s 1 core @ 2.5 Ghz (Python)
170 PS-fld code 12.23 % 19.03 % 10.53 % 0.25 s 1 core @ 2.5 Ghz (C/C++)
Y. Chen, H. Dai and Y. Ding: Pseudo-Stereo for Monocular 3D Object Detection in Autonomous Driving. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2022.
171 DD3Dv2 code 12.16 % 17.74 % 10.49 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
172 anonymity 12.00 % 18.98 % 10.21 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
173 OPA-3D code 11.01 % 17.14 % 9.94 % 0.04 s 1 core @ 3.5 Ghz (Python)
174 GCDR 10.92 % 15.65 % 9.86 % 0.28 s 1 core @ 2.5 Ghz (Python)
175 LT-M3OD 10.89 % 16.63 % 9.20 % 0.03 s 1 core @ 2.5 Ghz (Python)
176 ZongmuMono3d code 10.65 % 16.19 % 9.04 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
177 MonoDTR 10.59 % 16.66 % 9.00 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
K. Huang, T. Wu, H. Su and W. Hsu: MonoDTR: Monocular 3D Object Detection with Depth-Aware Transformer. CVPR 2022.
178 mono3d code 10.41 % 16.66 % 9.22 % TBD TBD
179 GUPNet code 10.37 % 15.62 % 8.79 % NA s 1 core @ 2.5 Ghz (Python + C/C++)
Y. Lu, X. Ma, L. Yang, T. Zhang, Y. Liu, Q. Chu, J. Yan and W. Ouyang: Geometry Uncertainty Projection Network for Monocular 3D Object Detection. arXiv preprint arXiv:2107.13774 2021.
180 CMKD* 10.28 % 16.03 % 8.85 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
181 Lite-FPN-GUPNet 10.08 % 15.73 % 8.52 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
182 MonoInsight 9.98 % 15.20 % 8.90 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
183 GPENet code 9.96 % 15.47 % 8.55 % 0.02 s GPU @ 2.5 Ghz (Python)
184 gupnet_se 9.85 % 14.65 % 8.32 % 0.03s 1 core @ 2.5 Ghz (C/C++)
185 SwinMono3D 9.82 % 14.55 % 8.31 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
186 HBD 9.66 % 15.26 % 8.17 % 0.12 s 1 core @ 2.5 Ghz (C/C++)
187 SCSTSV-MonoFlex 9.62 % 14.45 % 8.14 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
188 SARM3D 9.42 % 14.32 % 8.15 % 0.03 s GPU @ 2.5 Ghz (Python)
189 CaDDN code 9.41 % 14.72 % 8.17 % 0.63 s GPU @ 2.5 Ghz (Python)
C. Reading, A. Harakeh, J. Chae and S. Waslander: Categorical Depth Distribution Network for Monocular 3D Object Detection. CVPR 2021.
190 SGM3D 9.39 % 15.39 % 8.61 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
Z. Zhou, L. Du, X. Ye, Z. Zou, X. Tan, E. Ding, L. Zhang, X. Xue and J. Feng: SGM3D: Stereo Guided Monocular 3D Object Detection. 2021.
191 MonoGround 9.11 % 13.67 % 7.68 % 0.03 s 1 core @ 2.5 Ghz (Python)
192 K3D 9.06 % 14.56 % 7.59 % 0.3 s 1 core @ 2.5 Ghz (C/C++)
193 M3DGAF 8.93 % 13.42 % 7.58 % 0.07 s 1 core @ 2.5 Ghz (Python)
194 MonoFlex 8.91 % 13.26 % 7.38 % 0.03 s 1 core @ 2.5 Ghz (Python)
195 MonoFar 8.89 % 12.73 % 7.72 % 0.04 s 1 core @ 2.5 Ghz (Python)
196 SAIC_ADC_Mono3D code 8.87 % 13.92 % 7.55 % 50 s GPU @ 2.5 Ghz (Python)
197 MonoEdge 8.87 % 13.33 % 7.50 % 0.05 s GPU @ 2.5 Ghz (Python)
198 HomoLoss(monoflex) code 8.81 % 13.26 % 7.41 % 0.04 s 1 core @ 2.5 Ghz (Python)
J. Gu, B. Wu, L. Fan, J. Huang, S. Cao, Z. Xiang and X. Hua: Homography Loss for Monocular 3D Object Detection. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2022.
199 MonoCon code 8.73 % 13.55 % 7.83 % 0.02 s GPU @ 2.5 Ghz (Python)
T. Xianpeng Liu: Learning Auxiliary Monocular Contexts Helps Monocular 3D Object Detection. AAAI 2022.
200 MonoDDE 8.41 % 12.38 % 7.16 % 0.04 s 1 core @ 2.5 Ghz (Python)
Z. Li, Z. Qu, Y. Zhou, J. Liu, H. Wang and L. Jiang: Diversity Matters: Fully Exploiting Depth Clues for Reliable Monocular 3D Object Detection. CVPR 2022.
201 Mix-Teaching 8.40 % 12.34 % 7.06 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
202 ANM 8.28 % 12.83 % 7.59 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
203 MDSNet 8.18 % 12.05 % 7.03 % 0.07 s 1 core @ 2.5 Ghz (Python)
204 DCD code 8.08 % 11.76 % 6.61 % 1 s 1 core @ 2.5 Ghz (C/C++)
205 mono3d 7.95 % 11.89 % 6.75 % 0.03 s GPU @ 2.5 Ghz (Python)
206 MonoAug 7.94 % 12.66 % 6.64 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
207 LPCG-Monoflex 7.92 % 12.11 % 6.61 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
208 RefinedMPL 7.92 % 13.09 % 7.25 % 0.15 s GPU @ 2.5 Ghz (Python + C/C++)
J. Vianney, S. Aich and B. Liu: RefinedMPL: Refined Monocular PseudoLiDAR for 3D Object Detection in Autonomous Driving. arXiv preprint arXiv:1911.09712 2019.
209 Shape-Aware 7.65 % 11.69 % 6.35 % 0.05 s 1 core @ 2.5 Ghz (Python)
210 MonoRUn code 7.59 % 11.70 % 6.34 % 0.07 s GPU @ 2.5 Ghz (Python + C/C++)
H. Chen, Y. Huang, W. Tian, Z. Gao and L. Xiong: MonoRUn: Monocular 3D Object Detection by Reconstruction and Uncertainty Propagation. IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2021.
211 MonoEdge-Rotate 7.53 % 11.62 % 6.79 % 0.05 s GPU @ 2.5 Ghz (Python)
212 MonoFlex 7.36 % 10.36 % 6.29 % 0.03 s GPU @ 2.5 Ghz (Python)
Y. Zhang, J. Lu and J. Zhou: Objects are Different: Flexible Monocular 3D Object Detection. CVPR 2021.
213 MonoPair 7.04 % 10.99 % 6.29 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
Y. Chen, L. Tai, K. Sun and M. Li: MonoPair: Monocular 3D Object Detection Using Pairwise Spatial Relationships. The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2020.
214 monodle code 6.96 % 10.73 % 6.20 % 0.04 s GPU @ 2.5 Ghz (Python)
X. Ma, Y. Zhang, D. Xu, D. Zhou, S. Yi, H. Li and W. Ouyang: Delving into Localization Errors for Monocular 3D Object Detection. CVPR 2021 .
215 MonoAug 6.87 % 10.81 % 5.66 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
216 TopNet-DecayRate
This method makes use of Velodyne laser scans.
 6.59 % 8.78 % 6.25 % 92 ms NVIDIA GeForce 1080 Ti (tensorflow-gpu)
S. Wirges, T. Fischer, C. Stiller and J. Frias: Object Detection and Classification in Occupancy Grid Maps Using Deep Convolutional Networks. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
217 M3DSSD++ code 6.19 % 9.24 % 5.54 % 0.16s 1 core @ 2.5 Ghz (C/C++)
218 MDNet 6.18 % 9.48 % 5.63 % 0.2 s 1 core @ 2.5 Ghz (C/C++)
219 MK3D 6.15 % 8.76 % 5.14 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
220 Shift R-CNN (mono) code 5.66 % 8.58 % 4.49 % 0.25 s GPU @ 1.5 Ghz (Python)
A. Naiden, V. Paunescu, G. Kim, B. Jeon and M. Leordeanu: Shift R-CNN: Deep Monocular 3D Object Detection With Closed-form Geometric Constraints. ICIP 2019.
221 MM 5.63 % 9.20 % 4.78 % 1 s 1 core @ 2.5 Ghz (C/C++)
222 Aug3D-RPN 5.22 % 7.14 % 4.21 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
C. He, J. Huang, X. Hua and L. Zhang: Aug3D-RPN: Improving Monocular 3D Object Detection by Synthetic Images with Virtual Depth. 2021.
223 Lite-FPN 4.79 % 7.13 % 4.26 % 0.01 s 1 core @ 2.5 Ghz (C/C++)
224 COF3D 4.78 % 7.20 % 4.55 % 200 s 1 core @ 2.5 Ghz (Python + C/C++)
225 MAOLoss code 4.74 % 6.63 % 4.19 % 0.05 s 1 core @ 2.5 Ghz (Python)
226 TopNet-UncEst
This method makes use of Velodyne laser scans.
 4.60 % 6.88 % 3.79 % 0.09 s NVIDIA GeForce 1080 Ti (tensorflow-gpu)
S. Wirges, M. Braun, M. Lauer and C. Stiller: Capturing Object Detection Uncertainty in Multi-Layer Grid Maps. 2019.
227 MP-Mono 4.59 % 6.04 % 3.96 % 0.16 s GPU @ 2.5 Ghz (Python)
228 MonoPSR code 4.56 % 7.24 % 4.11 % 0.2 s GPU @ 3.5 Ghz (Python)
J. Ku*, A. Pon* and S. Waslander: Monocular 3D Object Detection Leveraging Accurate Proposals and Shape Reconstruction. CVPR 2019.
229 DFR-Net 4.52 % 6.66 % 3.71 % 0.18 s 1080 Ti (Pytorch)
Z. Zou, X. Ye, L. Du, X. Cheng, X. Tan, L. Zhang, J. Feng, X. Xue and E. Ding: The devil is in the task: Exploiting reciprocal appearance-localization features for monocular 3d object detection . ICCV 2021.
230 MoGDE 4.51 % 7.22 % 3.83 % 0.03 s GPU @ 2.5 Ghz (Python)
231 QD-3DT
This is an online method (no batch processing).
 code 4.23 % 6.62 % 3.39 % 0.03 s GPU @ 2.5 Ghz (Python)
H. Hu, Y. Yang, T. Fischer, F. Yu, T. Darrell and M. Sun: Monocular Quasi-Dense 3D Object Tracking. ArXiv:2103.07351 2021.
232 M3D-RPN code 4.05 % 5.65 % 3.29 % 0.16 s GPU @ 1.5 Ghz (Python)
G. Brazil and X. Liu: M3D-RPN: Monocular 3D Region Proposal Network for Object Detection . ICCV 2019 .
233 DDMP-3D 4.02 % 5.53 % 3.36 % 0.18 s 1 core @ 2.5 Ghz (Python)
L. Wang, L. Du, X. Ye, Y. Fu, G. Guo, X. Xue, J. Feng and L. Zhang: Depth-conditioned Dynamic Message Propagation for Monocular 3D Object Detection. CVPR 2020.
234 CMAN 3.96 % 5.24 % 3.18 % 0.15 s 1 core @ 2.5 Ghz (Python)
235 D4LCN code 3.86 % 5.06 % 3.59 % 0.2 s GPU @ 2.5 Ghz (Python + C/C++)
M. Ding, Y. Huo, H. Yi, Z. Wang, J. Shi, Z. Lu and P. Luo: Learning Depth-Guided Convolutions for Monocular 3D Object Detection. CVPR 2020.
236 RT3DStereo
This method uses stereo information.
 3.65 % 4.72 % 3.00 % 0.08 s GPU @ 2.5 Ghz (C/C++)
H. Königshof, N. Salscheider and C. Stiller: Realtime 3D Object Detection for Automated Driving Using Stereo Vision and Semantic Information. Proc. IEEE Intl. Conf. Intelligent Transportation Systems 2019.
237 MonoEF 3.05 % 4.61 % 2.85 % 0.03 s 1 core @ 2.5 Ghz (Python)
Y. Zhou, Y. He, H. Zhu, C. Wang, H. Li and Q. Jiang: Monocular 3D Object Detection: An Extrinsic Parameter Free Approach. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2021.
238 SS3D 2.09 % 2.48 % 1.61 % 48 ms Tesla V100 (Python)
E. Jörgensen, C. Zach and F. Kahl: Monocular 3D Object Detection and Box Fitting Trained End-to-End Using Intersection-over-Union Loss. CoRR 2019.
239 SparVox3D 2.05 % 2.90 % 1.69 % 0.05 s GPU @ 2.0 Ghz (Python)
E. Balatkan and F. Kıraç: Improving Regression Performance on Monocular 3D Object Detection Using Bin-Mixing and Sparse Voxel Data. 2021 6th International Conference on Computer Science and Engineering (UBMK) 2021.
240 CDTrack3D code 1.91 % 2.56 % 1.49 % 0.0106 s NVIDIA RTX 3090 GPU, i9 10850k CPU
241 PGD-FCOS3D code 1.88 % 2.82 % 1.54 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
T. Wang, X. Zhu, J. Pang and D. Lin: Probabilistic and Geometric Depth: Detecting Objects in Perspective. Conference on Robot Learning (CoRL) 2021.
242 EM code 1.25 % 1.18 % 0.89 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
243 mBoW
This method makes use of Velodyne laser scans.
 0.00 % 0.00 % 0.00 % 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.
Table as LaTeX | Only published Methods

Cyclist


Method Setting Code Moderate Easy Hard Runtime Environment
1 CasA++ 76.99 % 88.93 % 70.10 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
2 TED 76.95 % 89.54 % 70.31 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
3 CasA 75.74 % 88.99 % 68.47 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
4 ISE-RCNN-PV 75.40 % 86.08 % 68.58 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
5 ISE-RCNN 74.49 % 85.93 % 67.96 % 0.09 s 1 core @ 2.5 Ghz (Python + C/C++)
6 HMFI code 74.06 % 85.69 % 67.11 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
7 SGNet 73.88 % 88.03 % 66.84 % 0.09 s GPU @ 2.5 Ghz (Python)
8 SARFE 73.84 % 85.63 % 66.31 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
9 EQ-PVRCNN code 73.30 % 86.25 % 65.49 % 0.2 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Yang, L. Jiang, Y. Sun, B. Schiele and J. Jia: A Unified Query-based Paradigm for Point Cloud Understanding. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2022.
10 Semantical PVRCNN 73.14 % 86.75 % 64.87 % 0.07 s 1 core @ 2.5 Ghz (C/C++)
11 Anonymous 73.08 % 85.29 % 66.46 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
12 SPT 72.90 % 86.10 % 65.13 % 0.1 s GPU @ 2.5 Ghz (Python)
13 CAD 72.87 % 87.09 % 65.78 % 0.1 s GPU @ 2.5 Ghz (Python + C/C++)
14 DCAN-Second code 72.74 % 88.62 % 65.89 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
15 FS-Net
This method makes use of Velodyne laser scans.
 72.61 % 84.43 % 65.88 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
16 DSA-PV-RCNN
This method makes use of Velodyne laser scans.
 code 72.61 % 83.93 % 65.82 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
P. Bhattacharyya, C. Huang and K. Czarnecki: SA-Det3D: Self-Attention Based Context-Aware 3D Object Detection. 2021.
17 CAT-Det 72.51 % 85.35 % 65.55 % 0.3 s GPU @ 2.5 Ghz (Python + C/C++)
Y. Zhang, J. Chen and D. Huang: CAT-Det: Contrastively Augmented Transformer for Multi-modal 3D Object Detection. CVPR 2022.
18 Reprod-Two-Branch 72.16 % 87.50 % 64.41 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
19 IKT3D
This method makes use of Velodyne laser scans.
 72.12 % 83.56 % 64.12 % 0.05 s 1 core @ 2.5 Ghz (Python)
20 CFF-tv 72.02 % 86.54 % 64.25 % 1 s 1 core @ 2.5 Ghz (C/C++)
21 CFF-ep25 71.99 % 86.78 % 64.18 % 1 s 1 core @ 2.5 Ghz (C/C++)
22 VCRCNN 71.93 % 84.06 % 64.07 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
23 PV-RCNN++ code 71.86 % 84.60 % 63.84 % 0.06 s 1 core @ 2.5 Ghz (C/C++)
24 BtcDet
This method makes use of Velodyne laser scans.
 code 71.76 % 84.48 % 64.70 % 0.09 s GPU @ 2.5 Ghz (Python + C/C++)
Q. Xu, Y. Zhong and U. Neumann: Behind the Curtain: Learning Occluded Shapes for 3D Object Detection. Proceedings of the AAAI Conference on Artificial Intelligence 2022.
25 CF-ctdep-tv-ta 71.74 % 87.38 % 64.30 % 1 s 1 core @ 2.5 Ghz (C/C++)
26 cff-tv-v2-ep25 71.70 % 85.61 % 64.15 % 1 s 1 core @ 2.5 Ghz (C/C++)
27 PointPainting
This method makes use of Velodyne laser scans.
 71.54 % 83.91 % 62.97 % 0.4 s GPU @ 2.5 Ghz (Python + C/C++)
S. Vora, A. Lang, B. Helou and O. Beijbom: PointPainting: Sequential Fusion for 3D Object Detection. CVPR 2020.
28 CFF-tv-v2 71.53 % 85.70 % 63.77 % 1 s 1 core @ 2.5 Ghz (C/C++)
29 RangeIoUDet
This method makes use of Velodyne laser scans.
 71.49 % 85.99 % 63.62 % 0.02 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Liang, Z. Zhang, M. Zhang, X. Zhao and S. Pu: RangeIoUDet: Range Image Based Real-Time 3D Object Detector Optimized by Intersection Over Union. CVPR 2021.
30 IA-SSD (single) code 71.44 % 85.91 % 63.41 % 0.013 s 1 core @ 2.5 Ghz (C/C++)
Y. Zhang, Q. Hu, G. Xu, Y. Ma, J. Wan and Y. Guo: Not All Points Are Equal: Learning Highly Efficient Point-based Detectors for 3D LiDAR Point Clouds. CVPR 2022.
31 3SNet 71.44 % 84.55 % 64.79 % 0.07 s GPU @ 2.5 Ghz (Python)
32 PDV code 71.31 % 85.54 % 64.40 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
J. Hu, T. Kuai and S. Waslander: Point Density-Aware Voxels for LiDAR 3D Object Detection. CVPR 2022.
33 PE-RCVN 71.18 % 85.95 % 64.38 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
34 HVNet 71.17 % 83.97 % 63.65 % 0.03 s GPU @ 2.0 Ghz (Python)
M. Ye, S. Xu and T. Cao: HVNet: Hybrid Voxel Network for LiDAR Based 3D Object Detection. CVPR 2020.
35 M3DeTR code 70.89 % 85.03 % 63.14 % n/a s GPU @ 1.0 Ghz (Python)
T. Guan, J. Wang, S. Lan, R. Chandra, Z. Wu, L. Davis and D. Manocha: M3DeTR: Multi-representation, Multi- scale, Mutual-relation 3D Object Detection with Transformers. 2021.
36 USVLab BSAODet 70.85 % 85.28 % 64.09 % 0.04 s GPU @ 2.5 Ghz (Python + C/C++)
37 DDet 70.76 % 84.81 % 63.98 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
38 CenterFuse 70.59 % 86.53 % 62.18 % 0.059 sec/frame 2 x V100
39 TBD code 70.59 % 83.06 % 63.07 % 0.01 s 1 core @ 2.5 Ghz (C/C++)
40 AutoAlign 70.55 % 85.98 % 63.39 % 0.1 s 1 core @ 2.5 Ghz (Python)
41 MSADet 70.38 % 86.58 % 63.63 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
42 CZY 70.32 % 86.42 % 63.32 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
43 TCDVF 70.28 % 82.85 % 63.54 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
44 USVLab BSAODet (S) 70.24 % 84.38 % 63.24 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
45 MVMM code 70.17 % 81.84 % 63.84 % 0.04 s GPU @ 2.5 Ghz (Python + C/C++)
46 CF-ctdep-tv 70.16 % 86.31 % 62.63 % 1 s 1 core @ 2.5 Ghz (C/C++)
47 SPG_mini
This method makes use of Velodyne laser scans.
 code 70.09 % 82.66 % 63.61 % 0.09 s GPU @ 2.5 Ghz (Python)
Q. Xu, Y. Zhou, W. Wang, C. Qi and D. Anguelov: SPG: Unsupervised Domain Adaptation for 3D Object Detection via Semantic Point Generation. Proceedings of the IEEE conference on computer vision and pattern recognition (ICCV) 2021.
48 TBD 70.09 % 82.60 % 63.39 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
49 TBD 69.97 % 79.75 % 63.46 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
50 FV2P v2 69.82 % 86.88 % 63.09 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
51 KeyFuse2B 69.76 % 84.95 % 62.16 % 0.01 s 1 core @ 2.5 Ghz (C/C++)
52 CF-base-tv 69.49 % 84.12 % 61.85 % 1 s 1 core @ 2.5 Ghz (C/C++)
53 DGT-Det3D code 69.47 % 81.26 % 61.88 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
54 PVTr 69.46 % 84.62 % 62.33 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
55 ST-RCNN
This method makes use of Velodyne laser scans.
 69.42 % 80.69 % 62.63 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
56 IoU-2B 69.24 % 86.64 % 60.57 % 0.01 s 1 core @ 2.5 Ghz (C/C++)
57 DSASNet 69.12 % 82.32 % 62.68 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
58 TBD 69.09 % 82.53 % 62.57 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
59 JPVNet 69.07 % 83.46 % 62.73 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
60 MMLab PV-RCNN
This method makes use of Velodyne laser scans.
 code 68.89 % 82.49 % 62.41 % 0.08 s 1 core @ 2.5 Ghz (Python + C/C++)
S. Shi, C. Guo, L. Jiang, Z. Wang, J. Shi, X. Wang and H. Li: PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detection. CVPR 2020.
61 F-ConvNet
This method makes use of Velodyne laser scans.
 code 68.88 % 84.16 % 60.05 % 0.47 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Wang and K. Jia: Frustum ConvNet: Sliding Frustums to Aggregate Local Point-Wise Features for Amodal 3D Object Detection. IROS 2019.
62 FusionDetv2-v5 68.84 % 80.42 % 61.90 % 0.05 s 1 core @ 2.5 Ghz (Java + C/C++)
63 MMLab-PartA^2
This method makes use of Velodyne laser scans.
 code 68.73 % 83.43 % 61.85 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
S. Shi, Z. Wang, J. Shi, X. Wang and H. Li: From Points to Parts: 3D Object Detection from Point Cloud with Part-aware and Part-aggregation Network. IEEE Transactions on Pattern Analysis and Machine Intelligence 2020.
64 WGVRF 68.71 % 82.04 % 62.04 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
65 CF-cd-io-tv 68.52 % 83.71 % 60.12 % 1 s 1 core @ 2.5 Ghz (C/C++)
66 HotSpotNet 68.51 % 83.29 % 61.84 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
Q. Chen, L. Sun, Z. Wang, K. Jia and A. Yuille: object as hotspots. Proceedings of the European Conference on Computer Vision (ECCV) 2020.
67 TBD 68.33 % 85.17 % 61.38 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
68 NV2P-RCNN 68.31 % 78.63 % 61.08 % 0.1 s GPU @ 2.5 Ghz (Python)
69 FPV-SSD 68.15 % 80.32 % 60.51 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
70 ATT_SSD 68.14 % 81.49 % 61.31 % 0.01 s 1 core @ 2.5 Ghz (Python)
71 P2V-RCNN 68.06 % 81.09 % 60.73 % 0.1 s 2 cores @ 2.5 Ghz (Python)
J. Li, S. Luo, Z. Zhu, H. Dai, A. Krylov, Y. Ding and L. Shao: P2V-RCNN: Point to Voxel Feature Learning for 3D Object Detection from Point Clouds. IEEE Access 2021.
72 KPP3D code 67.97 % 81.23 % 60.72 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
73 H^23D R-CNN code 67.90 % 82.76 % 60.49 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
J. Deng, W. Zhou, Y. Zhang and H. Li: From Multi-View to Hollow-3D: Hallucinated Hollow-3D R-CNN for 3D Object Detection. IEEE Transactions on Circuits and Systems for Video Technology 2021.
74 Self-Calib Conv 67.73 % 82.11 % 60.57 % 0.01 s 1 core @ 2.5 Ghz (C/C++)
75 VPFNet code 67.66 % 80.83 % 61.36 % 0.2 s 1 core @ 2.5 Ghz (C/C++)
C. Wang, H. Chen and L. Fu: VPFNet: Voxel-Pixel Fusion Network for Multi-class 3D Object Detection. 2021.
76 3DSSD code 67.62 % 85.04 % 61.14 % 0.04 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Yang, Y. Sun, S. Liu and J. Jia: 3DSSD: Point-based 3D Single Stage Object Detector. CVPR 2020.
77 Fast-CLOCs 67.55 % 83.34 % 59.61 % 0.1 s GPU @ 2.5 Ghz (Python)
S. Pang, D. Morris and H. Radha: Fast-CLOCs: Fast Camera-LiDAR Object Candidates Fusion for 3D Object Detection. Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) 2022.
78 NV-RCNN 67.54 % 82.53 % 60.97 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
79 DGT-Det3D 67.44 % 80.73 % 60.71 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
80 cff-tv-t 67.41 % 85.91 % 60.15 % 1 s 1 core @ 2.5 Ghz (C/C++)
81 DVFENet 67.40 % 82.29 % 60.71 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
Y. He, G. Xia, Y. Luo, L. Su, Z. Zhang, W. Li and P. Wang: DVFENet: Dual-branch Voxel Feature Extraction Network for 3D Object Detection. Neurocomputing 2021.
82 FromVoxelToPoint code 67.36 % 82.68 % 59.15 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
J. Li, H. Dai, L. Shao and Y. Ding: From Voxel to Point: IoU-guided 3D Object Detection for Point Cloud with Voxel-to- Point Decoder. MM '21: The 29th ACM International Conference on Multimedia (ACM MM) 2021.
83 Point-GNN
This method makes use of Velodyne laser scans.
 code 67.28 % 81.17 % 59.67 % 0.6 s GPU @ 2.5 Ghz (Python)
W. Shi and R. Rajkumar: Point-GNN: Graph Neural Network for 3D Object Detection in a Point Cloud. CVPR 2020.
84 MMLab-PointRCNN
This method makes use of Velodyne laser scans.
 code 67.24 % 82.56 % 60.28 % 0.1 s GPU @ 2.5 Ghz (Python + C/C++)
S. Shi, X. Wang and H. Li: Pointrcnn: 3d object proposal generation and detection from point cloud. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2019.
85 STD code 67.23 % 81.36 % 59.35 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
Z. Yang, Y. Sun, S. Liu, X. Shen and J. Jia: STD: Sparse-to-Dense 3D Object Detector for Point Cloud. ICCV 2019.
86 SVGA-Net 66.82 % 81.25 % 59.37 % 0.03s 1 core @ 2.5 Ghz (Python + C/C++)
Q. He, Z. Wang, H. Zeng, Y. Zeng and Y. Liu: SVGA-Net: Sparse Voxel-Graph Attention Network for 3D Object Detection from Point Clouds. AAAI 2022.
87 KeyPoint-IoUHead 66.72 % 83.32 % 59.93 % 0.01 s 1 core @ 2.5 Ghz (C/C++)
88 S-AT GCN 66.71 % 78.53 % 60.19 % 0.02 s GPU @ 2.0 Ghz (Python)
L. Wang, C. Wang, X. Zhang, T. Lan and J. Li: S-AT GCN: Spatial-Attention Graph Convolution Network based Feature Enhancement for 3D Object Detection. CoRR 2021.
89 FusionDetv2-v4 66.54 % 82.50 % 59.78 % 0.06 s 1 core @ 2.5 Ghz (Python + C/C++)
90 T_PVRCNN_V2 66.49 % 80.88 % 58.51 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
91 cp-tv-kp-io-sc 66.40 % 82.88 % 58.53 % 1 s 1 core @ 2.5 Ghz (C/C++)
92 ARPNET 66.39 % 82.32 % 58.80 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
Y. Ye, C. Zhang and X. Hao: ARPNET: attention region proposal network for 3D object detection. Science China Information Sciences 2019.
93 IA-SSD (multi) code 66.29 % 81.30 % 59.58 % 0.014 s 1 core @ 2.5 Ghz (C/C++)
Y. Zhang, Q. Hu, G. Xu, Y. Ma, J. Wan and Y. Guo: Not All Points Are Equal: Learning Highly Efficient Point-based Detectors for 3D LiDAR Point Clouds. CVPR 2022.
94 T_PVRCNN 66.17 % 79.84 % 59.04 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
95 cp-tv 66.08 % 80.65 % 58.98 % 1 s 1 core @ 2.5 Ghz (C/C++)
96 MGAF-3DSSD code 66.00 % 83.03 % 57.57 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
J. Li, H. Dai, L. Shao and Y. Ding: Anchor-free 3D Single Stage Detector with Mask-Guided Attention for Point Cloud. MM '21: The 29th ACM International Conference on Multimedia (ACM MM) 2021.
97 AB3DMOT
This method makes use of Velodyne laser scans.
This is an online method (no batch processing).
 code 65.85 % 80.00 % 58.69 % 0.0047s 1 core @ 2.5 Ghz (Python)
X. Weng and K. Kitani: A Baseline for 3D Multi-Object Tracking. arXiv:1907.03961 2019.
98 VPN 65.60 % 82.20 % 58.96 % 0.06 s 1 core @ 2.5 Ghz (Python + C/C++)
99 PSA-Det3D 65.51 % 79.21 % 59.06 % 0.1 s GPU @ 2.5 Ghz (Python)
100 ZMMPP 65.23 % 77.62 % 58.08 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
101 cp-tv-kp 64.87 % 79.91 % 58.22 % 1 s 1 core @ 2.5 Ghz (C/C++)
102 TBD_BD code 64.60 % 82.19 % 58.01 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
103 Faraway-Frustum
This method makes use of Velodyne laser scans.
 code 64.54 % 79.65 % 57.84 % 0.1 s GPU @ 2.5 Ghz (Python)
H. Zhang, D. Yang, E. Yurtsever, K. Redmill and U. Ozguner: Faraway-frustum: Dealing with lidar sparsity for 3D object detection using fusion. 2021 IEEE International Intelligent Transportation Systems Conference (ITSC) 2021.
104 FusionDetv1 64.53 % 79.62 % 57.91 % 0.1 s 1 core @ 2.5 Ghz (Python + C/C++)
105 Dune-DCF-e11 64.52 % 82.14 % 57.40 % 1 s 1 core @ 2.5 Ghz (C/C++)
106 SRDL 64.52 % 79.64 % 57.90 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
ERROR: Wrong syntax in BIBTEX file.
107 Dune-DCF-e15 64.42 % 81.10 % 57.35 % 1 s 1 core @ 2.5 Ghz (C/C++)
108 City-CF-fixed 64.39 % 81.11 % 57.74 % 1 s 1 core @ 2.5 Ghz (C/C++)
109 CF-ctdep-train 64.33 % 81.02 % 56.17 % 1 s 1 core @ 2.5 Ghz (C/C++)
110 City-CF 64.25 % 81.33 % 57.01 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
111 CZY_3917 64.21 % 78.18 % 57.64 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
112 AGS-SSD[la] 64.19 % 76.17 % 57.48 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
113 SIF 64.13 % 79.32 % 57.38 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
P. An: SIF. Submitted to CVIU 2021.
114 AFTD 64.03 % 82.99 % 55.93 % 1 s 1 core @ 2.5 Ghz (Python + C/C++)
115 variance_point 63.90 % 78.49 % 56.51 % 0.05 s 1 core @ 2.5 Ghz (Python)
116 FusionDetv2-baseline 63.77 % 76.64 % 57.01 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
117 TANet code 63.77 % 79.16 % 56.21 % 0.035s GPU @ 2.5 Ghz (Python + C/C++)
Z. Liu, X. Zhao, T. Huang, R. Hu, Y. Zhou and X. Bai: TANet: Robust 3D Object Detection from Point Clouds with Triple Attention. AAAI 2020.
118 CF-base-train 63.63 % 80.31 % 55.56 % 1 s 1 core @ 2.5 Ghz (C/C++)
119 HS3D code 63.56 % 78.53 % 58.03 % 0.12 s 1 core @ 2.5 Ghz (Python + C/C++)
120 XView 63.06 % 81.32 % 56.65 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
L. Xie, G. Xu, D. Cai and X. He: X-view: Non-egocentric Multi-View 3D Object Detector. 2021.
121 EPNet++ 62.94 % 78.57 % 56.62 % 0.1 s GPU @ 2.5 Ghz (Python)
Z. Liu, H. tengteng, B. Li, X. Chen, X. Wang and X. Bai: EPNet++: Cascade Bi-directional Fusion for Multi-Modal 3D Object Detection. arXiv preprint arXiv:2112.11088 2021.
122 PointPillars
This method makes use of Velodyne laser scans.
 code 62.73 % 79.90 % 55.58 % 16 ms 1080ti GPU and Intel i7 CPU
A. Lang, S. Vora, H. Caesar, L. Zhou, J. Yang and O. Beijbom: PointPillars: Fast Encoders for Object Detection from Point Clouds. CVPR 2019.
123 Dune-DCF-e09 62.23 % 77.53 % 55.46 % 1 s 1 core @ 2.5 Ghz (C/C++)
124 Contrastive PP code 62.10 % 75.71 % 54.84 % 0.01 s 1 core @ 2.5 Ghz (Python)
125 CrazyTensor-CF 61.95 % 80.59 % 55.16 % 1 s 1 core @ 2.5 Ghz (C/C++)
126 PP-PCdet code 61.81 % 75.56 % 55.06 % 0.01 s 1 core @ 2.5 Ghz (Python)
127 LazyTorch-CP-Infer-O 61.40 % 76.40 % 54.73 % 1 s 1 core @ 2.5 Ghz (C/C++)
128 F-PointNet
This method makes use of Velodyne laser scans.
 code 61.37 % 77.26 % 53.78 % 0.17 s GPU @ 3.0 Ghz (Python)
C. Qi, W. Liu, C. Wu, H. Su and L. Guibas: Frustum PointNets for 3D Object Detection from RGB-D Data. arXiv preprint arXiv:1711.08488 2017.
129 CenterPoint (pcdet) 61.25 % 76.38 % 54.68 % 0.051 sec/frame 2 x V100
130 LazyTorch-CP-Small-P 61.07 % 76.37 % 54.73 % 1 s 1 core @ 2.5 Ghz (C/C++)
131 TBD 60.58 % 76.98 % 53.49 % 0.05 s 1 core @ 2.5 Ghz (C/C++)
132 TBD 60.58 % 76.98 % 53.49 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
133 epBRM
This method makes use of Velodyne laser scans.
 code 59.79 % 75.13 % 53.36 % 0.10 s 1 core @ 2.5 Ghz (C/C++)
K. Shin: Improving a Quality of 3D Object Detection by Spatial Transformation Mechanism. arXiv preprint arXiv:1910.04853 2019.
134 BirdNet+
This method makes use of Velodyne laser scans.
 code 59.58 % 70.84 % 54.20 % 0.11 s Titan Xp (PyTorch)
A. Barrera, J. Beltrán, C. Guindel, J. Iglesias and F. García: BirdNet+: Two-Stage 3D Object Detection in LiDAR through a Sparsity-Invariant Bird’s Eye View. IEEE Access 2021.
135 CrazyTensor-CP 59.54 % 75.40 % 53.21 % 1 s 1 core @ 2.5 Ghz (Python)
136 GT3D 58.64 % 76.60 % 52.07 % 0.1 s 1 core @ 2.5 Ghz (Python)
137 DMF
This method uses stereo information.
 57.99 % 71.92 % 51.55 % 0.2 s 1 core @ 2.5 Ghz (Python + C/C++)
X. J. Chen and W. Xu: Disparity-Based Multiscale Fusion Network for Transportation Detection. IEEE Transactions on Intelligent Transportation Systems 2022.
138 PointRGBNet 57.59 % 73.09 % 51.78 % 0.08 s 4 cores @ 2.5 Ghz (Python + C/C++)
P. Xie Desheng: Real-time Detection of 3D Objects Based on Multi-Sensor Information Fusion. Automotive Engineering 2022.
139 tbd 57.15 % 72.89 % 50.29 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
140 AVOD-FPN
This method makes use of Velodyne laser scans.
 code 57.12 % 69.39 % 51.09 % 0.1 s Titan X (Pascal)
J. Ku, M. Mozifian, J. Lee, A. Harakeh and S. Waslander: Joint 3D Proposal Generation and Object Detection from View Aggregation. IROS 2018.
141 PiFeNet 56.94 % 72.80 % 50.04 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
D. Le, H. Shi, H. Rezatofighi and J. Cai: PiFeNet: Pillar-Feature Network for Real-Time 3D Pedestrian Detection from Point Cloud. 2021.
142 CZY 56.71 % 70.64 % 50.35 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
143 SCNet
This method makes use of Velodyne laser scans.
 56.39 % 73.73 % 49.99 % 0.04 s GPU @ 3.0 Ghz (Python)
Z. Wang, H. Fu, L. Wang, L. Xiao and B. Dai: SCNet: Subdivision Coding Network for Object Detection Based on 3D Point Cloud. IEEE Access 2019.
144 PFF3D
This method makes use of Velodyne laser scans.
 code 55.71 % 72.67 % 49.58 % 0.05 s GPU @ 3.0 Ghz (Python + C/C++)
L. Wen and K. Jo: Fast and Accurate 3D Object Detection for Lidar-Camera-Based Autonomous Vehicles Using One Shared Voxel-Based Backbone. IEEE Access 2021.
145 MLOD
This method makes use of Velodyne laser scans.
 code 55.06 % 73.03 % 48.21 % 0.12 s GPU @ 1.5 Ghz (Python)
J. Deng and K. Czarnecki: MLOD: A multi-view 3D object detection based on robust feature fusion method. arXiv preprint arXiv:1909.04163 2019.
146 BirdNet+ (legacy)
This method makes use of Velodyne laser scans.
 code 52.15 % 72.45 % 46.57 % 0.1 s Titan Xp (PyTorch)
A. Barrera, C. Guindel, J. Beltrán and F. García: BirdNet+: End-to-End 3D Object Detection in LiDAR Bird’s Eye View. 2020 IEEE 23rd International Conference on Intelligent Transportation Systems (ITSC) 2020.
147 DSGN++
This method uses stereo information.
 code 49.37 % 68.29 % 43.79 % 0.2 s GeForce RTX 2080Ti
Y. Chen, S. Huang, S. Liu, B. Yu and J. Jia: DSGN++: Exploiting Visual-Spatial Relation for Stereo-based 3D Detectors. arXiv preprint arXiv:2204.03039 2022.
148 StereoDistill 48.37 % 69.46 % 42.69 % 0.4 s 1 core @ 2.5 Ghz (Python)
149 AVOD
This method makes use of Velodyne laser scans.
 code 48.15 % 64.11 % 42.37 % 0.08 s Titan X (pascal)
J. Ku, M. Mozifian, J. Lee, A. Harakeh and S. Waslander: Joint 3D Proposal Generation and Object Detection from View Aggregation. IROS 2018.
150 BirdNet
This method makes use of Velodyne laser scans.
 41.56 % 58.64 % 36.94 % 0.11 s Titan Xp (Caffe)
J. Beltrán, C. Guindel, F. Moreno, D. Cruzado, F. García and A. Escalera: BirdNet: A 3D Object Detection Framework from LiDAR Information. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
151 SparsePool code 40.74 % 56.52 % 36.68 % 0.13 s 8 cores @ 2.5 Ghz (Python)
Z. Wang, W. Zhan and M. Tomizuka: Fusing bird view lidar point cloud and front view camera image for deep object detection. arXiv preprint arXiv:1711.06703 2017.
152 MMLAB LIGA-Stereo
This method uses stereo information.
 code 40.60 % 58.95 % 35.27 % 0.4 s 1 core @ 2.5 Ghz (Python + C/C++)
X. Guo, S. Shi, X. Wang and H. Li: LIGA-Stereo: Learning LiDAR Geometry Aware Representations for Stereo-based 3D Detector. Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) 2021.
153 TopNet-Retina
This method makes use of Velodyne laser scans.
 36.83 % 47.48 % 33.58 % 52ms GeForce 1080Ti (tensorflow-gpu, v1.12)
S. Wirges, T. Fischer, C. Stiller and J. Frias: Object Detection and Classification in Occupancy Grid Maps Using Deep Convolutional Networks. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
154 CG-Stereo
This method uses stereo information.
 36.25 % 55.33 % 32.17 % 0.57 s GeForce RTX 2080 Ti
C. Li, J. Ku and S. Waslander: Confidence Guided Stereo 3D Object Detection with Split Depth Estimation. IROS 2020.
155 PS++ code 35.75 % 54.06 % 31.17 % PS++ s 1 core @ 2.5 Ghz (C/C++)
156 SparsePool code 35.24 % 43.55 % 30.15 % 0.13 s 8 cores @ 2.5 Ghz (Python)
Z. Wang, W. Zhan and M. Tomizuka: Fusing bird view lidar point cloud and front view camera image for deep object detection. arXiv preprint arXiv:1711.06703 2017.
157 PS code 32.16 % 49.23 % 27.73 % PS s 1 core @ 2.5 Ghz (C/C++)
158 Disp R-CNN (velo)
This method uses stereo information.
 code 27.04 % 44.19 % 23.58 % 0.387 s GPU @ 2.5 Ghz (Python + C/C++)
J. Sun, L. Chen, Y. Xie, S. Zhang, Q. Jiang, X. Zhou and H. Bao: Disp R-CNN: Stereo 3D Object Detection via Shape Prior Guided Instance Disparity Estimation. CVPR 2020.
159 Disp R-CNN
This method uses stereo information.
 code 27.04 % 44.19 % 23.58 % 0.387 s GPU @ 2.5 Ghz (Python + C/C++)
J. Sun, L. Chen, Y. Xie, S. Zhang, Q. Jiang, X. Zhou and H. Bao: Disp R-CNN: Stereo 3D Object Detection via Shape Prior Guided Instance Disparity Estimation. CVPR 2020.
160 AEC3D 26.17 % 36.57 % 25.21 % 18 ms GPU @ 2.5 Ghz (Python)
161 Complexer-YOLO
This method makes use of Velodyne laser scans.
 25.43 % 32.00 % 22.88 % 0.06 s GPU @ 3.5 Ghz (C/C++)
M. Simon, K. Amende, A. Kraus, J. Honer, T. Samann, H. Kaulbersch, S. Milz and H. Michael Gross: Complexer-YOLO: Real-Time 3D Object Detection and Tracking on Semantic Point Clouds. The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops 2019.
162 DSGN
This method uses stereo information.
 code 21.04 % 31.23 % 18.93 % 0.67 s NVIDIA Tesla V100
Y. Chen, S. Liu, X. Shen and J. Jia: DSGN: Deep Stereo Geometry Network for 3D Object Detection. CVPR 2020.
163 OC Stereo
This method uses stereo information.
 code 19.23 % 32.47 % 17.11 % 0.35 s 1 core @ 2.5 Ghz (Python + C/C++)
A. Pon, J. Ku, C. Li and S. Waslander: Object-Centric Stereo Matching for 3D Object Detection. ICRA 2020.
164 TopNet-DecayRate
This method makes use of Velodyne laser scans.
 16.00 % 23.02 % 13.24 % 92 ms NVIDIA GeForce 1080 Ti (tensorflow-gpu)
S. Wirges, T. Fischer, C. Stiller and J. Frias: Object Detection and Classification in Occupancy Grid Maps Using Deep Convolutional Networks. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
165 RT3D-GMP
This method uses stereo information.
 13.92 % 20.59 % 12.74 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
H. Königshof and C. Stiller: Learning-Based Shape Estimation with Grid Map Patches for Realtime 3D Object Detection for Automated Driving. 2020 IEEE 23rd International Conference on Intelligent Transportation Systems (ITSC) 2020.
166 TopNet-UncEst
This method makes use of Velodyne laser scans.
 9.18 % 12.31 % 8.14 % 0.09 s NVIDIA GeForce 1080 Ti (tensorflow-gpu)
S. Wirges, M. Braun, M. Lauer and C. Stiller: Capturing Object Detection Uncertainty in Multi-Layer Grid Maps. 2019.
167 ESGN
This method uses stereo information.
 9.02 % 15.78 % 7.96 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
168 CMKD* 8.15 % 14.66 % 7.23 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
169 anonymity 7.39 % 12.38 % 6.65 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
170 PS-fld code 7.29 % 12.80 % 6.05 % 0.25 s 1 core @ 2.5 Ghz (C/C++)
Y. Chen, H. Dai and Y. Ding: Pseudo-Stereo for Monocular 3D Object Detection in Autonomous Driving. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2022.
171 Anonymous 7.24 % 12.53 % 6.21 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
172 DD3Dv2 code 7.02 % 10.67 % 5.78 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
173 anonymity 6.84 % 10.90 % 5.72 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
174 mono3d code 6.52 % 11.40 % 5.19 % TBD TBD
175 TopNet-HighRes
This method makes use of Velodyne laser scans.
 6.48 % 9.99 % 6.76 % 101ms NVIDIA GeForce 1080 Ti (tensorflow-gpu)
S. Wirges, T. Fischer, C. Stiller and J. Frias: Object Detection and Classification in Occupancy Grid Maps Using Deep Convolutional Networks. 2018 21st International Conference on Intelligent Transportation Systems (ITSC) 2018.
176 MonoPSR code 5.78 % 9.87 % 4.57 % 0.2 s GPU @ 3.5 Ghz (Python)
J. Ku*, A. Pon* and S. Waslander: Monocular 3D Object Detection Leveraging Accurate Proposals and Shape Reconstruction. CVPR 2019.
177 DD3D code 5.69 % 9.20 % 5.20 % n/a s 1 core @ 2.5 Ghz (C/C++)
D. Park, R. Ambrus, V. Guizilini, J. Li and A. Gaidon: Is Pseudo-Lidar needed for Monocular 3D Object detection?. IEEE/CVF International Conference on Computer Vision (ICCV) .
178 LT-M3OD 5.53 % 9.17 % 4.84 % 0.03 s 1 core @ 2.5 Ghz (Python)
179 CaDDN code 5.38 % 9.67 % 4.75 % 0.63 s GPU @ 2.5 Ghz (Python)
C. Reading, A. Harakeh, J. Chae and S. Waslander: Categorical Depth Distribution Network for Monocular 3D Object Detection. CVPR 2021.
180 Mix-Teaching 5.36 % 8.56 % 4.62 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
181 MonoInsight 4.97 % 7.40 % 4.09 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
182 LPCG-Monoflex 4.90 % 8.14 % 3.86 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
183 MDNet 4.74 % 8.10 % 4.19 % 0.2 s 1 core @ 2.5 Ghz (C/C++)
184 DEPT 4.71 % 8.82 % 4.15 % 0.03 s 1 core @ 2.5 Ghz (Python)
185 Lite-FPN-GUPNet 4.70 % 7.67 % 4.61 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
186 ZongmuMono3d code 4.63 % 8.72 % 3.94 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
187 Shape-Aware 4.60 % 8.00 % 4.50 % 0.05 s 1 core @ 2.5 Ghz (Python)
188 SAIC_ADC_Mono3D code 4.55 % 7.90 % 3.73 % 50 s GPU @ 2.5 Ghz (Python)
189 SCSTSV-MonoFlex 4.50 % 7.40 % 3.66 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
190 MAOLoss code 4.49 % 7.28 % 3.60 % 0.05 s 1 core @ 2.5 Ghz (Python)
191 MonoDDE 4.36 % 6.68 % 3.76 % 0.04 s 1 core @ 2.5 Ghz (Python)
Z. Li, Z. Qu, Y. Zhou, J. Liu, H. Wang and L. Jiang: Diversity Matters: Fully Exploiting Depth Clues for Reliable Monocular 3D Object Detection. CVPR 2022.
192 GPENet code 4.28 % 7.06 % 3.68 % 0.02 s GPU @ 2.5 Ghz (Python)
193 MonoDTR 4.11 % 5.84 % 3.48 % 0.04 s 1 core @ 2.5 Ghz (C/C++)
K. Huang, T. Wu, H. Su and W. Hsu: MonoDTR: Monocular 3D Object Detection with Depth-Aware Transformer. CVPR 2022.
194 RT3DStereo
This method uses stereo information.
 4.10 % 7.03 % 3.88 % 0.08 s GPU @ 2.5 Ghz (C/C++)
H. Königshof, N. Salscheider and C. Stiller: Realtime 3D Object Detection for Automated Driving Using Stereo Vision and Semantic Information. Proc. IEEE Intl. Conf. Intelligent Transportation Systems 2019.
195 HomoLoss(monoflex) code 4.09 % 6.81 % 3.78 % 0.04 s 1 core @ 2.5 Ghz (Python)
J. Gu, B. Wu, L. Fan, J. Huang, S. Cao, Z. Xiang and X. Hua: Homography Loss for Monocular 3D Object Detection. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2022.
196 DFR-Net 4.00 % 5.99 % 3.95 % 0.18 s 1080 Ti (Pytorch)
Z. Zou, X. Ye, L. Du, X. Cheng, X. Tan, L. Zhang, J. Feng, X. Xue and E. Ding: The devil is in the task: Exploiting reciprocal appearance-localization features for monocular 3d object detection . ICCV 2021.
197 SARM3D 3.85 % 5.59 % 3.28 % 0.03 s GPU @ 2.5 Ghz (Python)
198 GUPNet code 3.85 % 6.94 % 3.64 % NA s 1 core @ 2.5 Ghz (Python + C/C++)
Y. Lu, X. Ma, L. Yang, T. Zhang, Y. Liu, Q. Chu, J. Yan and W. Ouyang: Geometry Uncertainty Projection Network for Monocular 3D Object Detection. arXiv preprint arXiv:2107.13774 2021.
199 MoGDE 3.76 % 6.04 % 3.09 % 0.03 s GPU @ 2.5 Ghz (Python)
200 OPA-3D code 3.75 % 6.01 % 3.56 % 0.04 s 1 core @ 3.5 Ghz (Python)
201 MonoAug 3.71 % 5.66 % 3.00 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
202 SGM3D 3.63 % 7.05 % 3.33 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
Z. Zhou, L. Du, X. Ye, Z. Zou, X. Tan, E. Ding, L. Zhang, X. Xue and J. Feng: SGM3D: Stereo Guided Monocular 3D Object Detection. 2021.
203 DCD code 3.62 % 5.84 % 3.33 % 1 s 1 core @ 2.5 Ghz (C/C++)
204 K3D 3.39 % 6.16 % 3.13 % 0.3 s 1 core @ 2.5 Ghz (C/C++)
205 MonoFar 3.33 % 5.27 % 2.90 % 0.04 s 1 core @ 2.5 Ghz (Python)
206 Aug3D-RPN 3.33 % 5.44 % 2.82 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
C. He, J. Huang, X. Hua and L. Zhang: Aug3D-RPN: Improving Monocular 3D Object Detection by Synthetic Images with Virtual Depth. 2021.
207 mono3d 3.30 % 5.84 % 2.68 % 0.03 s GPU @ 2.5 Ghz (Python)
208 monodle code 3.28 % 5.34 % 2.83 % 0.04 s GPU @ 2.5 Ghz (Python)
X. Ma, Y. Zhang, D. Xu, D. Zhou, S. Yi, H. Li and W. Ouyang: Delving into Localization Errors for Monocular 3D Object Detection. CVPR 2021 .
209 ANM 3.26 % 4.90 % 2.87 % 0.02 s 1 core @ 2.5 Ghz (C/C++)
210 MonoGround 3.22 % 4.81 % 2.75 % 0.03 s 1 core @ 2.5 Ghz (Python)
211 MDSNet 3.22 % 5.99 % 2.62 % 0.07 s 1 core @ 2.5 Ghz (Python)
212 DDMP-3D 3.14 % 4.92 % 2.44 % 0.18 s 1 core @ 2.5 Ghz (Python)
L. Wang, L. Du, X. Ye, Y. Fu, G. Guo, X. Xue, J. Feng and L. Zhang: Depth-conditioned Dynamic Message Propagation for Monocular 3D Object Detection. CVPR 2020.
213 M3DSSD++ code 3.14 % 5.73 % 3.03 % 0.16s 1 core @ 2.5 Ghz (C/C++)
214 QD-3DT
This is an online method (no batch processing).
 code 3.02 % 5.71 % 2.73 % 0.03 s GPU @ 2.5 Ghz (Python)
H. Hu, Y. Yang, T. Fischer, F. Yu, T. Darrell and M. Sun: Monocular Quasi-Dense 3D Object Tracking. ArXiv:2103.07351 2021.
215 M3DGAF 3.02 % 5.33 % 2.87 % 0.07 s 1 core @ 2.5 Ghz (Python)
216 MonoEdge-Rotate 3.01 % 5.36 % 2.83 % 0.05 s GPU @ 2.5 Ghz (Python)
217 MonoPair 2.87 % 4.76 % 2.42 % 0.06 s GPU @ 2.5 Ghz (Python + C/C++)
Y. Chen, L. Tai, K. Sun and M. Li: MonoPair: Monocular 3D Object Detection Using Pairwise Spatial Relationships. The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2020.
218 SwinMono3D 2.78 % 4.50 % 2.53 % 0.08 s 1 core @ 2.5 Ghz (C/C++)
219 MonoCon code 2.68 % 3.87 % 2.24 % 0.02 s GPU @ 2.5 Ghz (Python)
T. Xianpeng Liu: Learning Auxiliary Monocular Contexts Helps Monocular 3D Object Detection. AAAI 2022.
220 MonoFlex 2.67 % 4.41 % 2.50 % 0.03 s GPU @ 2.5 Ghz (Python)
Y. Zhang, J. Lu and J. Zhou: Objects are Different: Flexible Monocular 3D Object Detection. CVPR 2021.
221 gupnet_se 2.61 % 4.38 % 2.34 % 0.03s 1 core @ 2.5 Ghz (C/C++)
222 MK3D 2.55 % 4.17 % 2.24 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
223 MonoEdge 2.54 % 4.02 % 2.43 % 0.05 s GPU @ 2.5 Ghz (Python)
224 MonoFlex 2.51 % 4.36 % 2.38 % 0.03 s 1 core @ 2.5 Ghz (Python)
225 MonoAug 2.46 % 4.31 % 2.21 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
226 RefinedMPL 2.42 % 4.23 % 2.14 % 0.15 s GPU @ 2.5 Ghz (Python + C/C++)
J. Vianney, S. Aich and B. Liu: RefinedMPL: Refined Monocular PseudoLiDAR for 3D Object Detection in Autonomous Driving. arXiv preprint arXiv:1911.09712 2019.
227 GCDR 2.11 % 3.74 % 1.99 % 0.28 s 1 core @ 2.5 Ghz (Python)
228 SS3D 1.89 % 3.45 % 1.44 % 48 ms Tesla V100 (Python)
E. Jörgensen, C. Zach and F. Kahl: Monocular 3D Object Detection and Box Fitting Trained End-to-End Using Intersection-over-Union Loss. CoRR 2019.
229 D4LCN code 1.82 % 2.72 % 1.79 % 0.2 s GPU @ 2.5 Ghz (Python + C/C++)
M. Ding, Y. Huo, H. Yi, Z. Wang, J. Shi, Z. Lu and P. Luo: Learning Depth-Guided Convolutions for Monocular 3D Object Detection. CVPR 2020.
230 PGD-FCOS3D code 1.79 % 3.54 % 1.56 % 0.03 s 1 core @ 2.5 Ghz (C/C++)
T. Wang, X. Zhu, J. Pang and D. Lin: Probabilistic and Geometric Depth: Detecting Objects in Perspective. Conference on Robot Learning (CoRL) 2021.
231 MP-Mono 1.74 % 2.78 % 1.86 % 0.16 s GPU @ 2.5 Ghz (Python)
232 HBD 1.64 % 3.15 % 1.76 % 0.12 s 1 core @ 2.5 Ghz (C/C++)
233 COF3D 1.60 % 2.70 % 1.55 % 200 s 1 core @ 2.5 Ghz (Python + C/C++)
234 CMAN 1.48 % 1.76 % 1.17 % 0.15 s 1 core @ 2.5 Ghz (Python)
235 MonoEF 1.18 % 2.36 % 1.15 % 0.03 s 1 core @ 2.5 Ghz (Python)
Y. Zhou, Y. He, H. Zhu, C. Wang, H. Li and Q. Jiang: Monocular 3D Object Detection: An Extrinsic Parameter Free Approach. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2021.
236 M3D-RPN code 0.81 % 1.25 % 0.78 % 0.16 s GPU @ 1.5 Ghz (Python)
G. Brazil and X. Liu: M3D-RPN: Monocular 3D Region Proposal Network for Object Detection . ICCV 2019 .
237 MonoRUn code 0.73 % 1.14 % 0.66 % 0.07 s GPU @ 2.5 Ghz (Python + C/C++)
H. Chen, Y. Huang, W. Tian, Z. Gao and L. Xiong: MonoRUn: Monocular 3D Object Detection by Reconstruction and Uncertainty Propagation. IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2021.
238 Lite-FPN 0.44 % 0.52 % 0.27 % 0.01 s 1 core @ 2.5 Ghz (C/C++)
239 MM 0.40 % 0.71 % 0.35 % 1 s 1 core @ 2.5 Ghz (C/C++)
240 Shift R-CNN (mono) code 0.38 % 0.76 % 0.41 % 0.25 s GPU @ 1.5 Ghz (Python)
A. Naiden, V. Paunescu, G. Kim, B. Jeon and M. Leordeanu: Shift R-CNN: Deep Monocular 3D Object Detection With Closed-form Geometric Constraints. ICIP 2019.
241 CDTrack3D code 0.10 % 0.24 % 0.11 % 0.0106 s NVIDIA RTX 3090 GPU, i9 10850k CPU
242 mBoW
This method makes use of Velodyne laser scans.
 0.00 % 0.00 % 0.00 % 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.
Table as LaTeX | Only published Methods

Related Datasets

Citation

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}
}



eXTReMe Tracker