MOT17Det Results

Click on a measure to sort the table accordingly. See below for a more detailed description.


DetectorAPMODAMODPFAFTPFPFNPrecisionRecall
ACF
1.
0.32
18.172.12.837,31216,53977,25269.332.6
P. Dollar, R. Appel, S. Belongie, P. Perona. Fast Feature Pyramids for Object Detection. In TPAMI, 2014.
DPM
2.
0.61
31.275.87.178,00742,30836,55764.868.1
P. Felzenszwalb, R. Girshick, D. McAllester, D. Ramanan. Object Detection with Discriminatively Trained Part Based Models. In TPAMI, 2010.
FRCNN
3.
0.72
68.578.01.788,60110,08125,96389.877.3
S. Ren, K. He, R. Girshick, J. Sun. Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks. In NIPS, 2015.
HDGP
4.
0.45
42.176.41.355,6807,43658,88488.248.6
A. Garcia-Martin, R. Sanchez-Matilla, J. Martinez. Hierarchical detection of persons in groups. In Signal, Image and Video Processing, 2017.
KDNT
5.
0.89
67.180.14.8105,47328,6239,09178.792.1
F. Yu, W. Li, Q. Li, Y. Liu, X. Shi, J. Yan. POI: Multiple Object Tracking with High Performance Detection and Appearance Feature. In BMTT, SenseTime Group Limited, 2016.
MHD
6.
0.49
11.269.98.864,63751,80149,92755.556.4
Mobilenet-based Human Detection
SDP
7.
0.81
76.978.01.395,6997,59918,86592.683.5
F. Yang, W. Choi, Y. Lin. Exploit All the Layers: Fast and Accurate CNN Object Detector With Scale Dependent Pooling and Cascaded Rejection Classifiers. In The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016.
v3
8.
0.52
41.077.01.857,55410,55157,01084.550.2
Anonymous submission
VDet
9.
0.44
44.775.71.056,9805,76557,58490.849.7
Vitrociset Detection Algorithm
YLHD
10.
0.27
-37.065.614.040,59882,93973,96632.935.4
mobilenet based human detection
DetectorAPMODAMODPFAFTPFPFNPrecisionRecall
YLHDv2
11.
0.46
56.973.22.580,09314,93834,47184.369.9
https://arxiv.org/abs/1612.08242
yolov3
12.
0.77
62.974.73.794,14622,04720,41881.082.2
Anonymous submission
yolo_JTA
13.
0.62
54.371.52.375,84513,65838,71984.766.2
Anonymous submission
YTLAB
14.
0.89
76.780.22.8104,55516,68510,00986.291.3
Z. Cai, Q. Fan, R. Feris, N. Vasconcelos. A unified multi-scale deep convolutional neural network for fast object detection. In European Conference on Computer Vision, 2016.
ZIZOM
15.
0.81
72.079.82.295,41412,99019,13988.083.3
C. Lin, L. Jiwen, G. Wang, J. Zhou. Graininess-Aware Deep Feature Learning for Pedestrian Detection. In ECCV, 2018.

Benchmark Statistics

SequencesFramesBoxes
75919188076

Evaluation Measures

Lower is better. Higher is better.
Measure Better Perfect Description
AP higher 100 % Average Precision taken over a set of reference recall values (0:0.1:1)
MODA higher 100 % Multiple Object Detection Accuracy [1]. This measure combines false positives and missed targets.
MOTP higher 100 % Multiple Object Detection Precision [1]. The misalignment between the annotated and the predicted bounding boxes.
FAF lower 0 The average number of false alarms per frame.
TP higher #GT The total number of true positives.
FP lower 0 The total number of false positives.
FN lower 0 The total number of false negatives (missed targets).
Precision higher 100 % Ratio of TP / (TP+FP).
Recall higher 100 % Ratio of correct detections to total number of GT boxes.

Legend

Symbol Description
new This entry has been submitted or updated less than a week ago.

References:


[1] Stiefelhagen, R., Bernardin, K., Bowers, R., Garofolo, J.S., Mostefa, D. & Soundararajan, P. The CLEAR 2006 Evaluation. In CLEAR, 2006.