Submission details of MMPAT_CVPR21

Name MMPAT_CVPR21
Paper Link https://arxiv.org/abs/2105.14683
Code Link N/A
MOTA 0.316951
MOTP 0.202074
IDs 5742
False Positives 67171
False Negatives 580565
Input N/A
Runtime 0.1 s
Environment 1 GPU (Titan X)
Abstract Panoramic image detection and tracking using both image and point clouds. For tracking, we adopt the Cascade rcnn as our object detector. The backbone of the detector is ResNet 101. The detector is pre-trained on the COCO dataset and then fine tuned on the JRDB dataset. During training, we employ data augment methods such as mixup, random crop, multiscale training to augment the dataset. During inference, we employ horizontal flip and softnms techniques to generate more robust resutls. For tracking, we fuse both the 3D image and 3D point clouds for robust tracking.

Detailed results

Per-sequence results

sequence name MOTA MOTP IDs False Positives False Negatives
cubberly-auditorium-2019-04-22_1 45.04 21.23 111 829 6332
discovery-walk-2019-02-28_0 28.56 21.25 176 2382 6957
discovery-walk-2019-02-28_1 45.54 17.98 109 1200 6581
food-trucks-2019-02-12_0 49.22 18.20 238 1921 30392
gates-ai-lab-2019-04-17_0 39.98 15.22 134 2793 12573
gates-basement-elevators-2019-01-17_0 51.55 15.80 64 187 5759
gates-foyer-2019-01-17_0 58.25 15.37 54 1688 2511
gates-to-clark-2019-02-28_0 45.47 13.93 25 424 942
hewlett-class-2019-01-23_0 50.17 16.83 126 1117 6996
hewlett-class-2019-01-23_1 81.88 14.87 31 144 756
huang-2-2019-01-25_1 36.73 21.37 44 1090 3061
huang-intersection-2019-01-22_0 25.27 23.03 317 5989 28733
indoor-coupa-cafe-2019-02-06_0 31.62 20.52 305 3227 41324
lomita-serra-intersection-2019-01-30_0 13.92 19.70 83 4406 13238
meyer-green-2019-03-16_1 22.40 22.00 150 1992 16024
nvidia-aud-2019-01-25_0 40.38 22.36 227 2470 17053
nvidia-aud-2019-04-18_1 47.76 16.86 27 762 3382
nvidia-aud-2019-04-18_2 58.88 22.53 30 649 3508
outdoor-coupa-cafe-2019-02-06_0 19.94 21.01 282 4349 35524
quarry-road-2019-02-28_0 22.78 20.37 71 1430 3061
serra-street-2019-01-30_0 19.35 23.72 249 4984 29462
stlc-111-2019-04-19_1 63.71 13.91 38 490 1737
stlc-111-2019-04-19_2 58.10 14.61 24 419 1078
tressider-2019-03-16_2 17.98 23.15 230 4549 17084
tressider-2019-04-26_0 24.35 22.39 974 5436 84163
tressider-2019-04-26_1 30.34 18.64 770 4103 129246
tressider-2019-04-26_3 29.63 23.39 853 8141 73088
total 31.70 20.21 5742 67171 580565