search

Vchitect / Latte

Latte: Latent Diffusion Transformer for Video Generation.
Apache License 2.0
1.44k stars 147 forks source link

How to get preprocessed_ffs #82

Open zhang-haojie opened 1 month ago

zhang-haojie commented 1 month ago

Hello author, thank you very much for your work.

When I run train_with_img.py, I see that preprocessed_ffs in the dataset needs images and train_list.txt. I don't know how to process FFS to get these data. In addition, I use original in FFS instead of videos, is this correct?

Thank you!

maxin-cn commented 1 month ago

Hello author, thank you very much for your work.

When I run train_with_img.py, I see that preprocessed_ffs in the dataset needs images and train_list.txt. I don't know how to process FFS to get these data. In addition, I use original in FFS instead of videos, is this correct?

Thank you!

Thanks for your interest. This is the content of train_list.txt (video-name/frame.jpg): image

Sorry, I'm not sure I understand what you mean by original. Could you please provide more information?

zhang-haojie commented 1 month ago

Thank you for your reply. I have downloaded the FFS dataset now. I would like to know if there is any preprocessing code available to generate the preprocessed_ffs/train_list.txt and preprocessed_ffs/train/images/ in config. The FFS dataset I downloaded has three categories under train: altered, mask and original. I used original as the preprocessed_ffs/train/videos/in config.