Submission details of AB3DMOT

Name AB3DMOT
Paper Link https://arxiv.org/pdf/1907.03961.pdf
Code Link https://github.com/xinshuoweng/AB3DMOT
MOTA 0.193465
MOTP 0.420195
IDs 6177
False Positives 13664
False Negatives 777946
Input N/A
Runtime 0.01 s
Environment 1 GPU (Titan X)
Abstract 3D multi-object tracking (MOT) is an essential component technology for many real-time applications such as autonomous driving or assistive robotics. Recent work on 3D MOT tend to focus more on developing accurate systems giving less regard to computational cost and system complexity. In contrast, this work proposes a simple yet accurate real-time 3D MOT system. We use an off-the-shelf 3D object detector to obtain oriented 3D bounding boxes from the LiDAR point cloud. Then, a combination of 3D Kalman filter and Hungarian algorithm is used for state estimation and data association. Although our baseline system is a straightforward combination of standard methods, we obtain the state-of-the-art results. To evaluate our baseline system, we propose a new 3D MOT extension to the official KITTI 2D MOT evaluation along with a set of new metrics. Our proposed baseline method for 3D MOT establishes new state-of-the-art performance on 3D MOT for KITTI. Surprisingly, although our baseline system does not use any 2D data as input, we place 2nd on the official KITTI 2D MOT leaderboard. Also, our proposed 3D MOT method runs at a rate of 214.7 FPS, achieving the fastest speed among all modern MOT systems.

Detailed results

Per-sequence results

sequence name MOTA MOTP IDs False Positives False Negatives
cubberly-auditorium-2019-04-22_1 28.74 41.02 109 219 9203
discovery-walk-2019-02-28_0 23.64 40.66 82 116 10055
discovery-walk-2019-02-28_1 29.16 39.38 103 345 9886
food-trucks-2019-02-12_0 28.97 44.95 439 381 44983
gates-ai-lab-2019-04-17_0 39.09 41.27 248 1042 14493
gates-basement-elevators-2019-01-17_0 48.16 36.71 110 129 5959
gates-foyer-2019-01-17_0 53.57 44.04 85 458 4342
gates-to-clark-2019-02-28_0 44.47 39.77 27 65 1324
hewlett-class-2019-01-23_0 49.27 42.76 227 225 7976
hewlett-class-2019-01-23_1 77.48 38.44 52 77 1042
huang-2-2019-01-25_1 32.23 37.82 54 176 4263
huang-intersection-2019-01-22_0 12.91 41.36 232 2730 37757
indoor-coupa-cafe-2019-02-06_0 14.82 42.97 417 939 58665
lomita-serra-intersection-2019-01-30_0 23.85 42.37 79 96 15430
meyer-green-2019-03-16_1 13.63 38.43 80 146 19788
nvidia-aud-2019-01-25_0 21.20 45.04 203 514 26433
nvidia-aud-2019-04-18_1 38.61 44.74 49 170 4676
nvidia-aud-2019-04-18_2 35.03 39.41 75 117 6494
outdoor-coupa-cafe-2019-02-06_0 8.51 46.97 215 860 49440
quarry-road-2019-02-28_0 16.58 34.63 34 359 4974
serra-street-2019-01-30_0 12.88 40.67 157 552 37701
stlc-111-2019-04-19_1 55.78 39.88 67 207 2505
stlc-111-2019-04-19_2 54.84 38.08 44 46 1551
tressider-2019-03-16_2 19.86 38.10 94 68 20899
tressider-2019-04-26_0 8.84 43.69 902 1455 115774
tressider-2019-04-26_1 17.74 39.71 1098 331 160259
tressider-2019-04-26_3 14.24 46.18 895 1841 102074
total 19.35 42.02 6177 13664 777946