-
Notifications
You must be signed in to change notification settings - Fork 597
/
metafile.yml
78 lines (74 loc) · 2.58 KB
/
metafile.yml
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
Collections:
- Name: QDTrack
Metadata:
Training Data: MOT17, crowdhuman, TAO, LVIS
Training Techniques:
- SGD
Training Resources: 8x V100 GPUs
Architecture:
- ResNet
Paper:
URL: https://arxiv.org/pdf/2006.06664.pdf
Title: Quasi-Dense Similarity Learning for Multiple Object Tracking
README: configs/mot/qdtrack/README.md
Models:
- Name: qdtrack_faster-rcnn_r50_fpn_4e_mot17-private-half
In Collection: QDTrack
Config: configs/mot/qdtrack/qdtrack_faster-rcnn_r50_fpn_4e_mot17-private-half.py
Metadata:
Training Data: MOT17
Training Memory (GB): 5.83
Epochs: 4
Results:
- Task: Multi-object Tracking
Dataset: MOT17
Metrics:
HOTA: 57.1
MOTA: 68.2
IDF1: 68.5
Weights: https://download.openmmlab.com/mmtracking/mot/qdtrack/mot_dataset/qdtrack_faster-rcnn_r50_fpn_4e_mot17_20220315_145635-76f295ef.pth
- Name: qdtrack_faster-rcnn_r50_fpn_4e_crowdhuman_mot17-private-half
In Collection: QDTrack
Config: configs/mot/qdtrack/qdtrack_faster-rcnn_r50_fpn_4e_crowdhuman_mot17-private-half.py
Metadata:
Training Data: MOT17, crowdhuman
Training Memory (GB): 6.31
Epochs: 4
Results:
- Task: Multi-object Tracking
Dataset: MOT17
Metrics:
HOTA: 59.1
MOTA: 71.7
IDF1: 71.6
Weights: https://download.openmmlab.com/mmtracking/mot/qdtrack/mot_dataset/qdtrack_faster-rcnn_r50_fpn_4e_crowdhuman_mot17_20220315_163453-68899b0a.pth
- Name: qdtrack_faster-rcnn_r101_fpn_24e_lvis
In Collection: QDTrack
Config: configs/mot/qdtrack/qdtrack_faster-rcnn_r101_fpn_24e_lvis.py
Metadata:
Training Data: LVIS
Training Memory (GB): 9.78
Epochs: 24
Results:
- Task: Multi-object Tracking
Dataset: TAO
Metrics:
AP: 17.2
AP50: 28.6
AP75: 17.7
Weights: https://download.openmmlab.com/mmtracking/mot/qdtrack/tao_dataset/qdtrack_faster-rcnn_r101_fpn_24e_lvis_20220430_024513-88911daf.pth
- Name: qdtrack_faster-rcnn_r101_fpn_12e_tao
In Collection: QDTrack
Config: configs/mot/qdtrack/qdtrack_faster-rcnn_r101_fpn_12e_tao.py
Metadata:
Training Data: TAO
Training Memory (GB): 1.36
Epochs: 12
Results:
- Task: Multi-object Tracking
Dataset: TAO
Metrics:
Track AP: 11.0
Track AP50: 15.8
Track AP75: 6.1
Weights: https://download.openmmlab.com/mmtracking/mot/qdtrack/tao_dataset/qdtrack_faster-rcnn_r101_fpn_12e_tao_20220613_211934-7cbf4062.pth