Skip to content

Conversation

@Soonwang1988
Copy link

No description provided.

removed status in case of no face found , would have to change image processor which is getting used many other places as well
@chunyu-li
Copy link
Collaborator

Have you tested this on your computer to see if it reduces memory usage?

@nitinmukesh
Copy link

@Soonwang1988

What does this pipeline do. Does it improve the quality or...?

@Soonwang1988
Copy link
Author

@chunyu-li yes i have tested it , so let me tell you what was the issue , lets say you have 16 gb of ram and you have 4k footage each frame is around 10mb in memory you can only have 16*1024/10 = 1600 frames in memory so ,1600 frames means around 1 min , thats why it was crashing, now i have loaded only single frame at a time and saved it in folder

@nitinmukesh
Copy link

nitinmukesh commented Mar 25, 2025

Tnx @Soonwang1988

So how to use it?
python -m scripts.align_video --video_path "/home/admin/Documents/latent_sync_workspace/test.mp4"

Don't we have to pass
parser.add_argument("--dim", type=int, required=True)

Also is it only for square video?
width=args.dim,
height=args.dim,

Will this output new video which need to be used with LatentSync pipeline? Sorry not a developer so any help is appreciated.

@Soonwang1988
Copy link
Author

this would give us folder where we would have aligned files on, which we can train AE for face enhancement

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

Successfully merging this pull request may close these issues.

3 participants