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

reproduce this excellent work #113

Open
vivva opened this issue Apr 14, 2021 · 2 comments
Open

reproduce this excellent work #113

vivva opened this issue Apr 14, 2021 · 2 comments

Comments

@vivva
Copy link

vivva commented Apr 14, 2021

Hello!I am new to here . I want to know how to reproduce this excellent work.
First of all, I only downloaded UCF101 from TH14 official website. And I extracted the frame and optical flow to / home / win / pcharmprojects / datasets / thumbos14 / frame. But when I tried Python gen_ proposal_ list.py Thumbos14 / home / win / pycharmprojects / datasets / thumbos14 / frame, it appears keyerror: 'video'_ validation_ 0000354'

I guess it's because my data structure is not right? Maybe I need to download something else?Ask for help! Thank you very much!

@yjxiong
Copy link
Owner

yjxiong commented Apr 14, 2021

Hi @vivva

The experiments need the THUMOS14 dataset. It is different from the UCF101 dataset. UCF101 contains trimmed videos while THUMOS14 contains long untrimmed videos used for the temporal action detection task. Please obtain the download link from the official website.

@vivva
Copy link
Author

vivva commented Apr 14, 2021

@yjxiong
Hi ! I am so glad to see your reply! Because I see that the training set on TH14's official website is 20 types in ucf101, I downloaded UCF to train first. So should I train and test with Val and test data sets?

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

2 participants