lizz

Perhaps the system is old? This thread might be related https://github.com/boostorg/filesystem/issues/172

Does the file exist? What's the result of `ls -al /home/liding/project/temporal_action_detection/mmaction2/data/thumos14/videos/val/video_validation_0000033.mp4` and `ffprobe -i /home/liding/project/temporal_action_detection/mmaction2/data/thumos14/videos/val/video_validation_0000033.mp4`?

The video looks normal from the output of ffprobe. Could you pls attach this video for debugging purpose?

> The video looks normal from the output of ffprobe. Could you pls attach this video for debugging purpose?

I guess the most suspicious reason is the data itself. So pls attach the problematic video, otherwise it is impossible to debug :)

Let me check

Please post the full compilation log of denseflow. It might used the wrong opencv lib

The log looks correct

Great to hear that. Compilation are always troublesome. If I got some time later on, I will make the installation compilation free :D

n rgb frames have (n-1) flows. The decision is because we support any steps (even negative) between to frames. Say if you set step=-2, then you get (n-2) flows.