Finspire13 / CMCS-Temporal-Action-Localization

Completeness Modeling and Context Separation for Weakly Supervised Temporal Action Localization (CVPR2019)
MIT License
150 stars 17 forks source link

The tool to extract optical flow and RGB frames #23

Closed xzxedu closed 4 years ago

xzxedu commented 4 years ago

Hi,

May I ask how do you format the original raw videos into 25 fps and convert them into the optical flow and RGB frames? Did you use OpenCV to do this?

Besides, what is the _flowx and _flowy in your tools/get_flow_intensity.py? Did all videos format to 25 fps and 340x256 resolution before generate the static video clips?

Cheers

Finspire13 commented 4 years ago

Convert to 25 FPS: ffmpeg

Convert into the optical flow and RGB frames: https://github.com/yjxiong/temporal-segment-networks