Faceswap Forum

See viewtopic.php?f=7&t=571

Chances are you didn't include the right alignment file. Make sure that the alignment file you're using matches the video you're converting.

Hey y'all!

If you have a miniguide as to how you started conversion - drop me a hint - and Matt/Kyle and/or Bryan - don't start chewing at me - I'm seriously stuck!

/sergor

P.S. Have a good and safe weekend!

We don't have a guide.

What are you specifically stuck on. Convert should be relatively straight forward.

torzdf wrote: ↑Fri Apr 24, 2020 9:02 am

We don't have a guide.

What are you specifically stuck on. Convert should be relatively straight forward.

Well first thanks Matt for getting back to me so fast - secondly - I understand (Convert-tab) the Alignments for excluding/including faces also path to model.

Example: If you abide strictly to the paths indicated in the extract guide (which I have), I have Index A (faces, cleaned align.) and Index B (faces, cleaned align.) and the Trained model currently 768000 iterations (fed it 20000 assorted faces so it's nowhere near ready to go.

I just need to know where to point what - like being a teenager again! And my offer sent to your inbox still stands btw!

/sergor

Ok, so you have a video you want to swap faces for.

This video should have an associated (cleaned) alignments file. If you don't have one, go and create one now (following the standard Extract workflow).

The path to this video is entered as input
The path to the model folder is entered as model_dir

Everything else should be self explanatory.

I would add to select ffmpeg as the writer, versus the default OpenCV. Just fill in Input, Output, Alignment and Model fields, and don't use masks initially - that's all you need.

Hello.

I successfully converter before, although this time it's telling me that

Model could not be found in folder 'F:\deepfake\#3 model villan'. Exiting

About the only advice I can give you is to make sure that you're in the right folder. The correct folder will have a state file, 3+ .h5 files (exact number depends on various things including model chosen) and a logs folder.

Thank you for your reply, I was in the correct folder, although it refused to work. I did the old restart PC "fix" and it worked. Very odd.

Sorry, but we don't support haunted systems. Suggest an exorcist Glad that a reboot seems to have fixed it.

I am trying to do test converting but it doesn't work.

In Convert tab I use Input Dir: Video in which I want face to be changed to new one.
Output dir: this is where it gets saved
Alignments: Well I don't know, I tried both folders (alignments with face of person from original video, aswell as alignments with face of person that I want her face to be used in video. Whatever alignments file I select, it gives me wall of not found pictures.

Both folders are full of images and one alignments.fsa file.

in Model Dir: I select folder D:/workspace/Models where I can see many files.

In regards to alignments, for original face there was one alignments file, but for NEW face for video, I used there 5 different video sources, so I ended up with 5 alignments, which I merged together like tutorial said.

Assuming alignments files are broken or whatever (even though they work for training), can I just make new extraction of original video, make alignment file out of those images and use that?

ps. Is this a problem if you change names of images before making new alignment file? Just name change that does not involve images being replaced by new ones

ps. When I say No alignment found for xxx.png, I mean it goes for over 1000 of images, nothing is found. xxx is just replaced name for the sake of topic name

You MUST use an alignments file that matches the video you're trying to convert, NOT your training alignments.

bryanlyon wrote: ↑Tue May 12, 2020 7:23 pm

You MUST use an alignments file that matches the video you're trying to convert, NOT your training alignments.

So lets say I lost it, even though I never deleted anything, but can I just create new one? I mean I can extract everything from this video right now and have alignments file, is that correct?

edit: I did it and it works, thought models after 50k iterations still don't look good, only few are almost on spot, but most of them are blurry or even cause entire scene to be blurry, so I think I have to keep it running.

please put steep by steep of how convert, what is a aligments A and how do I create it? input Dir are the faces we want to add to the video?, Reference video is the video where faces will be replaced?

I put the file aligment.fsa en "Alignments" to convert but it says "no aligment found for 2020_000008 skipping

I honestly don't have time to write a convert guide at the moment (Documentation takes time, as does developing the software, as does the rest of my life).

We would be happy to sticky a user provided guide.

See the extract guide for how to generate an alignments file for convert, and why it's needed.

torzdf wrote: ↑Thu May 21, 2020 9:55 am

I honestly don't have time to write a convert guide at the moment (Documentation takes time, as does developing the software, as does the rest of my life).

We would be happy to sticky a user provided guide.

See the extract guide for how to generate an alignments file for convert, and why it's needed.

Se que ninguno de los desarrolladores del programa tiene tiempo para finalizar la guía, ¿cuanto debo pagarles y a que cuenta para que terminen la guía? yo he podido finalizar el proceso de conversión con éxito dos veces, pero intento volver a convertir ya con la PC mas entrenada y aun que hago todo igual no vuelve a salir... ¿cuanto pides por terminar la guia?

I sugest we all who want the finished guide pay 5 USD$ dolars to the creator of this program so he can finish the guide, who is agree? I am disposed to pay, if we are 10 persons we will give them 50 dolars to finish the guide, if we are 100 persons we will give them 500 dolars

Hello, so I've been training for about 10 hours, and has 33k iterations.

I had the source of the video cut into image each frame by another software and had the face extracted from the images I got by the faceswap software.

For the destination video, I got the face extracted by another software and just use that extracted faces when training.

Is this a problem when converting? I kept on getting the no alignments error.

Thank you.

I'm not sure I fully understand, but you need to generate an alignments file for the final video/images so that the convert process knows where to place the faces