Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Performance Benchmarks for video decoding backends #5

Open
iejMac opened this issue Jul 12, 2022 · 6 comments
Open

Performance Benchmarks for video decoding backends #5

iejMac opened this issue Jul 12, 2022 · 6 comments

Comments

@iejMac
Copy link
Owner

iejMac commented Jul 12, 2022

No description provided.

@iejMac iejMac changed the title benchmarking method for readers benchmarking method for frame readers Jul 12, 2022
@iejMac
Copy link
Owner Author

iejMac commented Jul 13, 2022

WIP - #11

@iejMac
Copy link
Owner Author

iejMac commented Jul 13, 2022

Perhaps we should have a more representative benchmark dataset. Currently as an initial mp4 dataset I use one class of Kinetics700 train videos which is about 1000 videos of length ~25 seconds. The benchmark might not be accurate for dataset with a longer average length

@iejMac
Copy link
Owner Author

iejMac commented Jul 13, 2022

Initial benchmark of cv2 on 6 CPU machine using 6 threads:
eff_cv2_50_224_6

htop output:
Screen Shot 2022-07-13 at 7 31 37 PM

@iejMac
Copy link
Owner Author

iejMac commented Jul 13, 2022

Initial benchmark of ffmpeg on 6 CPU machine using 6 threads:
eff_ffmpeg_50_224_6

htop output:
Screen Shot 2022-07-13 at 7 34 09 PM

@iejMac
Copy link
Owner Author

iejMac commented Jul 18, 2022

@iejMac
Copy link
Owner Author

iejMac commented Jul 21, 2022

Improved cv2 read with #13:
test

@iejMac iejMac changed the title benchmarking method for frame readers Performance Benchmarks for video decoding backends Jul 31, 2022
@iejMac iejMac pinned this issue Jul 31, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant