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

On deblurring after video synthesis #11

Open
Explore008 opened this issue Aug 8, 2023 · 1 comment
Open

On deblurring after video synthesis #11

Explore008 opened this issue Aug 8, 2023 · 1 comment

Comments

@Explore008
Copy link

Hello, author. Your work in this field is excellent. I am currently studying a work about video-driven two-dimensional human image movement, and finally a video is generated, but the frame rate is low and the artifacts are very serious. I want to introduce your work into my article and optimize it directly in the final generated video. How should I call it?

@XiangZ-0
Copy link
Owner

XiangZ-0 commented Aug 9, 2023

Hi, thanks for your interest in our work. Assuming you already got the synthesized videos and the corresponding events, training EVDI on them is the same as training on regular videos. All you need to do is first prepare your data via 'Prepare_data.py' and then run 'Train.py' to optimize your EVDI model. Please see the 'Train' section in README.md for detailed instructions. Hope this helps!

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