simplexsigil2 commited on
Commit
f1b7744
·
verified ·
1 Parent(s): fe6210f

Upload folder using huggingface_hub

Browse files
README.md CHANGED
@@ -1,15 +1,3 @@
1
- ---
2
- license: cc-by-nc-4.0
3
- task_categories:
4
- - video-classification
5
- language:
6
- - en
7
- tags:
8
- - medical
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
 
15
  [![License: CC BY-NC-SA 4.0](https://img.shields.io/badge/License-CC%20BY--NC--SA%204.0-lightgrey.svg)](https://creativecommons.org/licenses/by-nc-sa/4.0/)
@@ -35,14 +23,14 @@ OmniFall addresses three critical limitations in current fall detection research
35
  This benchmark includes annotations for the following datasets:
36
 
37
  1. **CMDFall** (7h 7m) - Multi-subject, multi-view dataset
38
- 2. **UP Fall** (4h 35m) - Multiple subjects with two camera views
39
  3. **Le2i** (47m) - Multiple room settings
40
- 4. **GMDCSA24** (21m) - Multiple room settings
41
- 5. **CAUCAFall** (16m)
42
- 6. **OCCU** (14m) - Two camera views
43
- 7. **EDF** (13m) - Two camera views
44
  8. **MCFD** (12m) - Single subject, eight camera views
45
- 9. **OOPS-Fall** - Curated subset of genuine fall accidents from the OOPS dataset
46
 
47
  ## Structure
48
 
@@ -184,7 +172,7 @@ wild_test_data = dataset["test"]
184
 
185
  ## Citation
186
 
187
- If you use OmniFall in your research, please cite our paper (will be updated soon):
188
 
189
  ```bibtex
190
  @inproceedings{omnifall2025,
@@ -203,4 +191,4 @@ The original video data belongs to their respective owners and should be obtaine
203
 
204
  ## Contact
205
 
206
- For questions about the dataset, please contact [david.schneider@kit.edu].
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  # OmniFall: A Unified Benchmark for Staged-to-Wild Fall Detection
2
 
3
  [![License: CC BY-NC-SA 4.0](https://img.shields.io/badge/License-CC%20BY--NC--SA%204.0-lightgrey.svg)](https://creativecommons.org/licenses/by-nc-sa/4.0/)
 
23
  This benchmark includes annotations for the following datasets:
24
 
25
  1. **CMDFall** (7h 7m) - Multi-subject, multi-view dataset
26
+ 2. **UP Fall** (4h 35m) - Multiple subjects with varied camera views
27
  3. **Le2i** (47m) - Multiple room settings
28
+ 4. **GMDCSA24** (21m) - Varied lighting conditions
29
+ 5. **CAUCAFall** (16m) - Includes day/night recordings
30
+ 6. **OCCU** (14m) - Depth camera recordings
31
+ 7. **EDF** (13m) - Depth camera recordings
32
  8. **MCFD** (12m) - Single subject, eight camera views
33
+ 9. **OOPS-Fall** - Curated subset of genuine accidents from the OOPS dataset
34
 
35
  ## Structure
36
 
 
172
 
173
  ## Citation
174
 
175
+ If you use OmniFall in your research, please cite our paper:
176
 
177
  ```bibtex
178
  @inproceedings{omnifall2025,
 
191
 
192
  ## Contact
193
 
194
+ For questions about the dataset, please contact [david.schneider@kit.edu].
splits/cs/mcfd/test.csv CHANGED
@@ -0,0 +1 @@
 
 
1
+ path
splits/cs/mcfd/val.csv CHANGED
@@ -0,0 +1 @@
 
 
1
+ path
splits/cv/caucafall/test.csv CHANGED
@@ -0,0 +1 @@
 
 
1
+ path
splits/cv/caucafall/val.csv CHANGED
@@ -0,0 +1 @@
 
 
1
+ path