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

folder "cropped_videos_ant" for Hand Movement DataLoader #10

Open
Marchetz opened this issue Aug 2, 2022 · 3 comments
Open

folder "cropped_videos_ant" for Hand Movement DataLoader #10

Marchetz opened this issue Aug 2, 2022 · 3 comments

Comments

@Marchetz
Copy link

Marchetz commented Aug 2, 2022

Hi,
In the Ego4D Hand Movement Prediction Baseline, the procedure needs rescaled input videos in the folder "cropped_videos_ant".
Are these videos generated from the code of this baseline or do they need to be downloaded?
I downloaded the dataset with ego4d --output_directory="~/ego4d_data" --datasets annotations clips --metadata --benchmarks FHO
but it seems that these scaled video are not there.
Moreover, it seems that the baseline don't have the script to generate the rescaled videos.

Thanks

@antoninofurnari
Copy link
Collaborator

Hi, thank you for opining the issue and apologies if there has been no reply yet. Maybe @VJWQ can help?

@VJWQ
Copy link
Contributor

VJWQ commented Aug 21, 2022

Hi @Marchetz , we hope our user can choose their desired resolution and encourage them to explore the ego4d dataset freely instead of being restricted by the limited clips we used to build this benchmark, so we didn't provide rescaled videos. You can always use ffmepg and the annotation file we provided to construct the cropped clips that fit your model design best.

@kurshakuz
Copy link

Hello @VJWQ! I am curious if you already have the script that can used directly on the dataset to get the benchmark results as stated in the paper. Do you mind sharing it? Thanks!

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

4 participants