Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/kiyoon/video_datasets_api
Video datasets (EPIC-Kitchens, Something-Something, Diving48, HMDB-51, UCF-101) tools and API
https://github.com/kiyoon/video_datasets_api
Last synced: about 1 month ago
JSON representation
Video datasets (EPIC-Kitchens, Something-Something, Diving48, HMDB-51, UCF-101) tools and API
- Host: GitHub
- URL: https://github.com/kiyoon/video_datasets_api
- Owner: kiyoon
- Created: 2019-12-03T14:30:03.000Z (almost 5 years ago)
- Default Branch: master
- Last Pushed: 2022-11-21T03:42:17.000Z (almost 2 years ago)
- Last Synced: 2023-03-03T12:27:23.037Z (over 1 year ago)
- Language: Python
- Homepage:
- Size: 282 KB
- Stars: 4
- Watchers: 3
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# video_datasets_api
Video datasets (HMDB-51, EPIC-Kitchens-55/100, Something-Something-V1/V2, etc.) tools and API**Key features**
- Extract videos into frames
- If mjpeg dataset, just extract jpegs losslessly. Faster and better quality.
- Extract optical flows using multiple GPUs
- Definitions without having to load annotations (`NUM_CLASSES`)
- Read annotations (e.g. labels, class key strings)# Installation
You can use the tools as is. For the API, it is easier if you install.
`python setup.py develop`Then, you can import the package.
```python
from video_datasets_api.something_something_v1 import NUM_CLASSES
print(NUM_CLASSES) # 174
```