Hugging Face's logo

Datasets:
simplexsigil2
/
omnifall

Tasks:
Video Classification
Modalities:
Tabular
Text
Formats:
csv
Languages:
English
Size:
10K - 100K
ArXiv:
Tags:
medical
Libraries:
Datasets
Dask
License:
Dataset card Data Studio Files
xet
Community
1
simplexsigil2 commited on
Commit
1b96b67
·
verified ·
1 Parent(s): 22b7152

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +244 -0
README.md CHANGED
@@ -9,6 +9,250 @@ tags:
9
  pretty_name: 'OmniFall: A Unified Benchmark for Staged-to-Wild Fall Detection'
10
  size_categories:
11
  - 10K<n<100K
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
12
  ---
13
  # OmniFall: A Unified Benchmark for Staged-to-Wild Fall Detection
14
 
 
9
  pretty_name: 'OmniFall: A Unified Benchmark for Staged-to-Wild Fall Detection'
10
  size_categories:
11
  - 10K<n<100K
12
+
13
+ configs:
14
+ - config_name: default
15
+ data_files:
16
+ - split: train
17
+ path: "splits/cs/*/train.csv"
18
+ - split: validation
19
+ path: "splits/cs/*/val.csv"
20
+ - split: test
21
+ path: "splits/cs/*/test.csv"
22
+ description: "Cross-subject split configuration for all datasets"
23
+
24
+ # Cross-subject (cs) and Cross-validation (cv) base configurations
25
+ - config_name: cv
26
+ data_files:
27
+ - split: train
28
+ path: "splits/cv/*/train.csv"
29
+ - split: validation
30
+ path: "splits/cv/*/val.csv"
31
+ - split: test
32
+ path: "splits/cv/*/test.csv"
33
+ description: "Cross-validation split configuration for all datasets"
34
+
35
+ - config_name: cs
36
+ data_files:
37
+ - split: train
38
+ path: "splits/cs/*/train.csv"
39
+ - split: validation
40
+ path: "splits/cs/*/val.csv"
41
+ - split: test
42
+ path: "splits/cs/*/test.csv"
43
+ description: "Cross-subject split configuration for all datasets"
44
+
45
+ # New multi-dataset splits for staged/wild settings
46
+ - config_name: cs-staged
47
+ data_files:
48
+ - split: train
49
+ path:
50
+ - "splits/cs/caucafall/train.csv"
51
+ - "splits/cs/cmdfall/train.csv"
52
+ - "splits/cs/edf/train.csv"
53
+ - "splits/cs/gmdcsa24/train.csv"
54
+ - "splits/cs/le2i/train.csv"
55
+ - "splits/cs/mcfd/train.csv"
56
+ - "splits/cs/occu/train.csv"
57
+ - "splits/cs/up_fall/train.csv"
58
+ - split: validation
59
+ path:
60
+ - "splits/cs/caucafall/val.csv"
61
+ - "splits/cs/cmdfall/val.csv"
62
+ - "splits/cs/edf/val.csv"
63
+ - "splits/cs/gmdcsa24/val.csv"
64
+ - "splits/cs/le2i/val.csv"
65
+ - "splits/cs/mcfd/val.csv"
66
+ - "splits/cs/occu/val.csv"
67
+ - "splits/cs/up_fall/val.csv"
68
+ - split: test
69
+ path:
70
+ - "splits/cs/caucafall/test.csv"
71
+ - "splits/cs/cmdfall/test.csv"
72
+ - "splits/cs/edf/test.csv"
73
+ - "splits/cs/gmdcsa24/test.csv"
74
+ - "splits/cs/le2i/test.csv"
75
+ - "splits/cs/mcfd/test.csv"
76
+ - "splits/cs/occu/test.csv"
77
+ - "splits/cs/up_fall/test.csv"
78
+ description: "Cross-subject splits on staged datasets only (excluding OOPS)"
79
+
80
+ - config_name: cv-staged
81
+ data_files:
82
+ - split: train
83
+ path:
84
+ - "splits/cv/caucafall/train.csv"
85
+ - "splits/cv/cmdfall/train.csv"
86
+ - "splits/cv/edf/train.csv"
87
+ - "splits/cv/gmdcsa24/train.csv"
88
+ - "splits/cv/le2i/train.csv"
89
+ - "splits/cv/mcfd/train.csv"
90
+ - "splits/cv/occu/train.csv"
91
+ - "splits/cv/up_fall/train.csv"
92
+ - split: validation
93
+ path:
94
+ - "splits/cv/caucafall/val.csv"
95
+ - "splits/cv/cmdfall/val.csv"
96
+ - "splits/cv/edf/val.csv"
97
+ - "splits/cv/gmdcsa24/val.csv"
98
+ - "splits/cv/le2i/val.csv"
99
+ - "splits/cv/mcfd/val.csv"
100
+ - "splits/cv/occu/val.csv"
101
+ - "splits/cv/up_fall/val.csv"
102
+ - split: test
103
+ path:
104
+ - "splits/cv/caucafall/test.csv"
105
+ - "splits/cv/cmdfall/test.csv"
106
+ - "splits/cv/edf/test.csv"
107
+ - "splits/cv/gmdcsa24/test.csv"
108
+ - "splits/cv/le2i/test.csv"
109
+ - "splits/cv/mcfd/test.csv"
110
+ - "splits/cv/occu/test.csv"
111
+ - "splits/cv/up_fall/test.csv"
112
+ description: "Cross-validation splits on staged datasets only (excluding OOPS)"
113
+
114
+ - config_name: cs-staged-wild
115
+ data_files:
116
+ - split: train
117
+ path:
118
+ - "splits/cs/caucafall/train.csv"
119
+ - "splits/cs/cmdfall/train.csv"
120
+ - "splits/cs/edf/train.csv"
121
+ - "splits/cs/gmdcsa24/train.csv"
122
+ - "splits/cs/le2i/train.csv"
123
+ - "splits/cs/mcfd/train.csv"
124
+ - "splits/cs/occu/train.csv"
125
+ - "splits/cs/up_fall/train.csv"
126
+ - split: validation
127
+ path:
128
+ - "splits/cs/caucafall/val.csv"
129
+ - "splits/cs/cmdfall/val.csv"
130
+ - "splits/cs/edf/val.csv"
131
+ - "splits/cs/gmdcsa24/val.csv"
132
+ - "splits/cs/le2i/val.csv"
133
+ - "splits/cs/mcfd/val.csv"
134
+ - "splits/cs/occu/val.csv"
135
+ - "splits/cs/up_fall/val.csv"
136
+ - split: test
137
+ path: "splits/cs/OOPS/test.csv"
138
+ description: "Train and validate on staged datasets, test on wild (OOPS) dataset with cross-subject splits"
139
+
140
+ - config_name: cv-staged-wild
141
+ data_files:
142
+ - split: train
143
+ path:
144
+ - "splits/cv/caucafall/train.csv"
145
+ - "splits/cv/cmdfall/train.csv"
146
+ - "splits/cv/edf/train.csv"
147
+ - "splits/cv/gmdcsa24/train.csv"
148
+ - "splits/cv/le2i/train.csv"
149
+ - "splits/cv/mcfd/train.csv"
150
+ - "splits/cv/occu/train.csv"
151
+ - "splits/cv/up_fall/train.csv"
152
+ - split: validation
153
+ path:
154
+ - "splits/cv/caucafall/val.csv"
155
+ - "splits/cv/cmdfall/val.csv"
156
+ - "splits/cv/edf/val.csv"
157
+ - "splits/cv/gmdcsa24/val.csv"
158
+ - "splits/cv/le2i/val.csv"
159
+ - "splits/cv/mcfd/val.csv"
160
+ - "splits/cv/occu/val.csv"
161
+ - "splits/cv/up_fall/val.csv"
162
+ - split: test
163
+ path: "splits/cv/OOPS/test.csv"
164
+ description: "Train and validate on staged datasets, test on wild (OOPS) dataset with cross-validation splits"
165
+
166
+ # Individual dataset configurations
167
+ - config_name: OOPS
168
+ data_files:
169
+ - split: train
170
+ path: "splits/cs/OOPS/train.csv"
171
+ - split: validation
172
+ path: "splits/cs/OOPS/val.csv"
173
+ - split: test
174
+ path: "splits/cs/OOPS/test.csv"
175
+ description: "OOPS dataset with cross-subject splits"
176
+
177
+ - config_name: caucafall
178
+ data_files:
179
+ - split: train
180
+ path: "splits/cs/caucafall/train.csv"
181
+ - split: validation
182
+ path: "splits/cs/caucafall/val.csv"
183
+ - split: test
184
+ path: "splits/cs/caucafall/test.csv"
185
+ description: "CaucaFall dataset with cross-subject splits"
186
+
187
+ - config_name: cmdfall
188
+ data_files:
189
+ - split: train
190
+ path: "splits/cs/cmdfall/train.csv"
191
+ - split: validation
192
+ path: "splits/cs/cmdfall/val.csv"
193
+ - split: test
194
+ path: "splits/cs/cmdfall/test.csv"
195
+ description: "CMDFall dataset with cross-subject splits"
196
+
197
+ - config_name: edf
198
+ data_files:
199
+ - split: train
200
+ path: "splits/cs/edf/train.csv"
201
+ - split: validation
202
+ path: "splits/cs/edf/val.csv"
203
+ - split: test
204
+ path: "splits/cs/edf/test.csv"
205
+ description: "EDF dataset with cross-subject splits"
206
+
207
+ - config_name: gmdcsa24
208
+ data_files:
209
+ - split: train
210
+ path: "splits/cs/gmdcsa24/train.csv"
211
+ - split: validation
212
+ path: "splits/cs/gmdcsa24/val.csv"
213
+ - split: test
214
+ path: "splits/cs/gmdcsa24/test.csv"
215
+ description: "GMDCSA24 dataset with cross-subject splits"
216
+
217
+ - config_name: le2i
218
+ data_files:
219
+ - split: train
220
+ path: "splits/cs/le2i/train.csv"
221
+ - split: validation
222
+ path: "splits/cs/le2i/val.csv"
223
+ - split: test
224
+ path: "splits/cs/le2i/test.csv"
225
+ description: "LE2I dataset with cross-subject splits"
226
+
227
+ - config_name: mcfd
228
+ data_files:
229
+ - split: train
230
+ path: "splits/cs/mcfd/train.csv"
231
+ - split: validation
232
+ path: "splits/cs/mcfd/val.csv"
233
+ - split: test
234
+ path: "splits/cs/mcfd/test.csv"
235
+ description: "MCFD dataset with cross-subject splits"
236
+
237
+ - config_name: occu
238
+ data_files:
239
+ - split: train
240
+ path: "splits/cs/occu/train.csv"
241
+ - split: validation
242
+ path: "splits/cs/occu/val.csv"
243
+ - split: test
244
+ path: "splits/cs/occu/test.csv"
245
+ description: "OCCU dataset with cross-subject splits"
246
+
247
+ - config_name: up_fall
248
+ data_files:
249
+ - split: train
250
+ path: "splits/cs/up_fall/train.csv"
251
+ - split: validation
252
+ path: "splits/cs/up_fall/val.csv"
253
+ - split: test
254
+ path: "splits/cs/up_fall/test.csv"
255
+ description: "UP-Fall dataset with cross-subject splits"
256
  ---
257
  # OmniFall: A Unified Benchmark for Staged-to-Wild Fall Detection
258