Open ghileng opened 1 month ago
when i made use of IPAdapted FaceID to finsih video to animation,I found that some images had been resized and could not be detected by insightface. And if I do not use this plugin or just run in a short time with this plugin,this workflow can run normally
Hello, when I was testing different frames, I found that in a certain frame the video character blocked the face and reported an error. Then I made a small modification to the code of IPAdapterPlus.py as above and it can run normally, but I don’t know Is this the optimal solution for video-to-painting applications? but it may solve some of the problems in video-to-painting applications.
it's a bit more complicated than that. basically the whole attention patching has to be skipped if no faces are found. I'll think about it
Hi Matteo, this would also be incredibly useful for me as I do not have to prune off the entire branch of code if a face doesn't exist