Skip to content
This repository has been archived by the owner on Sep 3, 2024. It is now read-only.

Improve create_dataset, video_to_slomo #60

Open
wants to merge 2 commits into
base: master
Choose a base branch
from

Conversation

tobbez
Copy link

@tobbez tobbez commented Oct 12, 2019

  • In both: Use subprocess.check_call instead of os.system. This means even file names with special characters (e.g. ") will work correctly.
  • In create_dataset: Increase the number of digits used in frame file names. The previous number would cause frames to appear out of order when using them in training if a video used for the data set had more than 9999 frames.

This removes issues caused by file names containing special characters.
The previous length of 4 digits would cause frames out of order if you
used a video with more than 9999 frames, e.g. a 240 fps longer than ~42
seconds.

The new length of 9 should be high enough, with good margin (allowing
for 48+ days of 240 fps video).
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant