Skip to content

GitLab

Switch branch/tag
Find file Normal viewHistoryPermalink
metafile.yml 10 KB
Newer Older
Sugon_ldc's avatar
add model TSM  
Sugon_ldc committed
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
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237 
Collections:
- Name: I3D
  README: configs/recognition/i3d/README.md
  Paper:
    URL: https://arxiv.org/abs/1705.07750
    Title: Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset
Models:
- Config: configs/recognition/i3d/i3d_r50_32x2x1_100e_kinetics400_rgb.py
  In Collection: I3D
  Metadata:
    Architecture: ResNet50
    Batch Size: 8
    Epochs: 100
    FLOPs: 43564040192
    Parameters: 28043472
    Pretrained: ImageNet
    Resolution: 340x256
    Training Data: Kinetics-400
    Training Resources: 8 GPUs
  Modality: RGB
  Name: i3d_r50_32x2x1_100e_kinetics400_rgb
  Results:
  - Dataset: Kinetics-400
    Metrics:
      Top 1 Accuracy: 72.68
      Top 5 Accuracy: 90.78
    Task: Action Recognition
  Training Json Log: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_r50_32x2x1_100e_kinetics400_rgb/20200614_060456.log.json
  Training Log: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_r50_32x2x1_100e_kinetics400_rgb/20200614_060456.log
  Weights: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_r50_32x2x1_100e_kinetics400_rgb/i3d_r50_32x2x1_100e_kinetics400_rgb_20200614-c25ef9a4.pth
- Config: configs/recognition/i3d/i3d_r50_32x2x1_100e_kinetics400_rgb.py
  In Collection: I3D
  Metadata:
    Architecture: ResNet50
    Batch Size: 8
    Epochs: 100
    FLOPs: 43564040192
    Parameters: 28043472
    Pretrained: ImageNet
    Resolution: short-side 256
    Training Data: Kinetics-400
    Training Resources: 8 GPUs
  Modality: RGB
  Name: i3d_r50_32x2x1_100e_kinetics400_rgb
  Results:
  - Dataset: Kinetics-400
    Metrics:
      Top 1 Accuracy: 73.27
      Top 5 Accuracy: 90.92
    Task: Action Recognition
  Training Json Log: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_r50_256p_32x2x1_100e_kinetics400_rgb/20200725_031555.log.json
  Training Log: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_r50_256p_32x2x1_100e_kinetics400_rgb/20200725_031555.log
  Weights: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_r50_256p_32x2x1_100e_kinetics400_rgb/i3d_r50_256p_32x2x1_100e_kinetics400_rgb_20200801-7d9f44de.pth
- Config: configs/recognition/i3d/i3d_r50_video_32x2x1_100e_kinetics400_rgb.py
  In Collection: I3D
  Metadata:
    Architecture: ResNet50
    Batch Size: 8
    Epochs: 100
    FLOPs: 43564040192
    Parameters: 28043472
    Pretrained: ImageNet
    Resolution: short-side 256p
    Training Data: Kinetics-400
    Training Resources: 8 GPUs
  Modality: RGB
  Name: i3d_r50_video_32x2x1_100e_kinetics400_rgb
  Results:
  - Dataset: Kinetics-400
    Metrics:
      Top 1 Accuracy: 72.85
      Top 5 Accuracy: 90.75
    Task: Action Recognition
  Training Json Log: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_r50_video_32x2x1_100e_kinetics400_rgb/20200706_143014.log.json
  Training Log: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_r50_video_32x2x1_100e_kinetics400_rgb/20200706_143014.log
  Weights: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_r50_video_32x2x1_100e_kinetics400_rgb/i3d_r50_video_32x2x1_100e_kinetics400_rgb_20200826-e31c6f52.pth
- Config: configs/recognition/i3d/i3d_r50_dense_32x2x1_100e_kinetics400_rgb.py
  In Collection: I3D
  Metadata:
    Architecture: ResNet50
    Batch Size: 8
    Epochs: 100
    FLOPs: 43564040192
    Parameters: 28043472
    Pretrained: ImageNet
    Resolution: 340x256
    Training Data: Kinetics-400
    Training Resources: 16 GPUs
  Modality: RGB
  Name: i3d_r50_dense_32x2x1_100e_kinetics400_rgb
  Results:
  - Dataset: Kinetics-400
    Metrics:
      Top 1 Accuracy: 72.77
      Top 5 Accuracy: 90.57
    Task: Action Recognition
  Training Json Log: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_r50_dense_32x2x1_100e_kinetics400_rgb/20200616_230011.log.json
  Training Log: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_r50_dense_32x2x1_100e_kinetics400_rgb/20200616_230011.log
  Weights: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_r50_dense_32x2x1_100e_kinetics400_rgb/i3d_r50_dense_32x2x1_100e_kinetics400_rgb_20200616-2bbb4361.pth
- Config: configs/recognition/i3d/i3d_r50_dense_32x2x1_100e_kinetics400_rgb.py
  In Collection: I3D
  Metadata:
    Architecture: ResNet50
    Batch Size: 8
    Epochs: 100
    FLOPs: 43564040192
    Parameters: 28043472
    Pretrained: ImageNet
    Resolution: short-side 256
    Training Data: Kinetics-400
    Training Resources: 8 GPUs
  Modality: RGB
  Name: i3d_r50_dense_32x2x1_100e_kinetics400_rgb
  Results:
  - Dataset: Kinetics-400
    Metrics:
      Top 1 Accuracy: 73.48
      Top 5 Accuracy: 91.0
    Task: Action Recognition
  Training Json Log: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_r50_dense_256p_32x2x1_100e_kinetics400_rgb/20200725_031604.log.json
  Training Log: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_r50_dense_256p_32x2x1_100e_kinetics400_rgb/20200725_031604.log
  Weights: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_r50_dense_256p_32x2x1_100e_kinetics400_rgb/i3d_r50_dense_256p_32x2x1_100e_kinetics400_rgb_20200725-24eb54cc.pth
- Config: configs/recognition/i3d/i3d_r50_lazy_32x2x1_100e_kinetics400_rgb.py
  In Collection: I3D
  Metadata:
    Architecture: ResNet50
    Batch Size: 8
    Epochs: 100
    FLOPs: 43564040192
    Parameters: 28043472
    Pretrained: ImageNet
    Resolution: 340x256
    Training Data: Kinetics-400
    Training Resources: 8 GPUs
  Modality: RGB
  Name: i3d_r50_lazy_32x2x1_100e_kinetics400_rgb
  Results:
  - Dataset: Kinetics-400
    Metrics:
      Top 1 Accuracy: 72.32
      Top 5 Accuracy: 90.72
    Task: Action Recognition
  Training Json Log: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_r50_fast_32x2x1_100e_kinetics400_rgb/20200612_233836.log.json
  Training Log: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_r50_fast_32x2x1_100e_kinetics400_rgb/20200612_233836.log
  Weights: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_r50_fast_32x2x1_100e_kinetics400_rgb/i3d_r50_fast_32x2x1_100e_kinetics400_rgb_20200612-000e4d2a.pth
- Config: configs/recognition/i3d/i3d_r50_lazy_32x2x1_100e_kinetics400_rgb.py
  In Collection: I3D
  Metadata:
    Architecture: ResNet50
    Batch Size: 8
    Epochs: 100
    FLOPs: 43564040192
    Parameters: 28043472
    Pretrained: ImageNet
    Resolution: short-side 256
    Training Data: Kinetics-400
    Training Resources: 8 GPUs
  Modality: RGB
  Name: i3d_r50_lazy_32x2x1_100e_kinetics400_rgb
  Results:
  - Dataset: Kinetics-400
    Metrics:
      Top 1 Accuracy: 73.24
      Top 5 Accuracy: 90.99
    Task: Action Recognition
  Training Json Log: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_r50_fast_256p_32x2x1_100e_kinetics400_rgb/20200725_031457.log.json
  Training Log: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_r50_fast_256p_32x2x1_100e_kinetics400_rgb/20200725_031457.log
  Weights: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_r50_fast_256p_32x2x1_100e_kinetics400_rgb/i3d_r50_fast_256p_32x2x1_100e_kinetics400_rgb_20200817-4e90d1d5.pth
- Config: configs/recognition/i3d/i3d_nl_embedded_gaussian_r50_32x2x1_100e_kinetics400_rgb.py
  In Collection: I3D
  Metadata:
    Architecture: ResNet50
    Batch Size: 8
    Epochs: 100
    FLOPs: 54334488576
    Parameters: 35397840
    Pretrained: ImageNet
    Resolution: short-side 256p
    Training Data: Kinetics-400
    Training Resources: 32 GPUs
  Modality: RGB
  Name: i3d_nl_embedded_gaussian_r50_32x2x1_100e_kinetics400_rgb
  Results:
  - Dataset: Kinetics-400
    Metrics:
      Top 1 Accuracy: 74.71
      Top 5 Accuracy: 91.81
    Task: Action Recognition
  Training Json Log: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_nl_embedded_gaussian_r50_32x2x1_100e_kinetics400_rgb/20200813_034054.log.json
  Training Log: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_nl_embedded_gaussian_r50_32x2x1_100e_kinetics400_rgb/20200813_034054.log
  Weights: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_nl_embedded_gaussian_r50_32x2x1_100e_kinetics400_rgb/i3d_nl_embedded_gaussian_r50_32x2x1_100e_kinetics400_rgb_20200813-6e6aef1b.pth
- Config: configs/recognition/i3d/i3d_nl_gaussian_r50_32x2x1_100e_kinetics400_rgb.py
  In Collection: I3D
  Metadata:
    Architecture: ResNet50
    Batch Size: 8
    Epochs: 100
    FLOPs: 48962109440
    Parameters: 31723728
    Pretrained: ImageNet
    Resolution: short-side 256p
    Training Data: Kinetics-400
    Training Resources: 32 GPUs
  Modality: RGB
  Name: i3d_nl_gaussian_r50_32x2x1_100e_kinetics400_rgb
  Results:
  - Dataset: Kinetics-400
    Metrics:
      Top 1 Accuracy: 73.37
      Top 5 Accuracy: 91.26
    Task: Action Recognition
  Training Json Log: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_nl_gaussian_r50_32x2x1_100e_kinetics400_rgb/20200813_034909.log.json
  Training Log: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_nl_gaussian_r50_32x2x1_100e_kinetics400_rgb/20200813_034909.log
  Weights: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_nl_gaussian_r50_32x2x1_100e_kinetics400_rgb/i3d_nl_gaussian_r50_32x2x1_100e_kinetics400_rgb_20200815-17f84aa2.pth
- Config: configs/recognition/i3d/i3d_nl_dot_product_r50_32x2x1_100e_kinetics400_rgb.py
  In Collection: I3D
  Metadata:
    Architecture: ResNet50
    Batch Size: 8
    Epochs: 100
    FLOPs: 54334488576
    Parameters: 35397840
    Pretrained: ImageNet
    Resolution: short-side 256p
    Training Data: Kinetics-400
    Training Resources: 32 GPUs
  Modality: RGB
  Name: i3d_nl_dot_product_r50_32x2x1_100e_kinetics400_rgb
  Results:
  - Dataset: Kinetics-400
    Metrics:
      Top 1 Accuracy: 73.92
      Top 5 Accuracy: 91.59
    Task: Action Recognition
  Training Json Log: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_nl_dot_product_r50_32x2x1_100e_kinetics400_rgb/20200814_044208.log.json
  Training Log: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_nl_dot_product_r50_32x2x1_100e_kinetics400_rgb/20200814_044208.log
  Weights: https://download.openmmlab.com/mmaction/recognition/i3d/i3d_nl_dot_product_r50_32x2x1_100e_kinetics400_rgb/i3d_nl_dot_product_r50_32x2x1_100e_kinetics400_rgb_20200814-7c30d5bb.pth