MrDeepFakes Forums

Some content may not be available to Guests. Consider registering an account to enjoy unrestricted access to guides, support and tools

  • We are looking for community members who are intested in helping out. See our HELP WANTED post.

Openfaceswap no faces

jespo1352

DF Vagrant
Can't seem to pull faces from videos, image sets, or images pulled from videos. It gets everything, just doesn't get faces. Tried different videos, sizes, names. I have Cuda 9.0, cuDNN 7.05 (with files copied to Cuda directory, GTX 750 Nvidia, win 10, and Microsoft visual 2015. 

Any assistance would be appreciated

I Get this info below along with sample image set 


C:\OpenFaceSwap\python\python-3.6.3.amd64\lib\site-packages\h5py\__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.
  from ._conv import register_converters as _register_converters
Using TensorFlow backend.
Input Directory: C:\OpenFaceSwap\imgA
Output Directory: C:\OpenFaceSwap\alignA
Filter: filter.jpg
Using json serializer
Starting, this may take a while...
Loading Extract from Extract_Align plugin...
100%|████████████████████████████████████████████████████████████████████████████████| 215/215 [01:33<00:00,  2.29it/s]
Alignments filepath: C:\OpenFaceSwap\imgA\alignments.json
Writing alignments to: C:\OpenFaceSwap\imgA\alignments.json
-------------------------
Images found:        215
Faces detected:      0
-------------------------
Done!
Press any key to continue . . .
 

Pocketspeed

DF Admirer
Verified Video Creator
jespo1352 said:
Can't seem to pull faces from videos, image sets, or images pulled from videos. It gets everything, just doesn't get faces. Tried different videos, sizes, names. I have Cuda 9.0, cuDNN 7.05 (with files copied to Cuda directory, GTX 750 Nvidia, win 10, and Microsoft visual 2015. 

Any assistance would be appreciated

I Get this info below along with sample image set 


C:\OpenFaceSwap\python\python-3.6.3.amd64\lib\site-packages\h5py\__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.
  from ._conv import register_converters as _register_converters
Using TensorFlow backend.
Input Directory: C:\OpenFaceSwap\imgA
Output Directory: C:\OpenFaceSwap\alignA
Filter: filter.jpg
Using json serializer
Starting, this may take a while...
Loading Extract from Extract_Align plugin...
100%|████████████████████████████████████████████████████████████████████████████████| 215/215 [01:33<00:00,  2.29it/s]
Alignments filepath: C:\OpenFaceSwap\imgA\alignments.json
Writing alignments to: C:\OpenFaceSwap\imgA\alignments.json
-------------------------
Images found:        215
Faces detected:      0
-------------------------
Done!
Press any key to continue . . .


First thing I would check is Nvidia driver updates.
 

Pocketspeed

DF Admirer
Verified Video Creator
jespo1352 said:
Thank you. I actually did need some driver updates but still no faces are pulled.


dpfks said:
Try without a filter image

How do I do this? I don't see anything in the settings



When you run OpenFaceSwap, click the gear icon to the right of the Faces A box. Make sure that the filter box is unchecked.  Then do the same for the Faces B box.

Just a guess, but with a GTX 750, you might be stuck with running the LowMem version. It should be included in the OpenFaceSwap files.  Open your OpenFaceSwap location. Find "faceswap_lowmem". Copy the files inside. Go back to your OpenFaceSwap folder and paste (overwrite).  It should overwrite your Lib, Plugins, and Scripts folders, as well as your faceswap.py.  Close everything out.  Restart your PC.  Try again.  Alternatively, you can edit the text file.  To load the low memory faceswap packages, edit the openfaceswapconfig.txt file to point to the appropriate paths. This will normally only involve inserting a “_exp” or “_lowmem” in the appropriate paths.

If Fakeapp 2.2 is giving you trouble, download this Mega link:  https://mega.nz/#!HSIDgQCC!fjYJtkTepeX0hXWz3Qxp_EAp1TdD_u3gae_MX5h9ELE

Then reinstall, including all steps and dependencies. (CUDA, Visual Studio, everything.)

Let us know what you find.
 
I have did all of this....but it still shows the same error.....can anyone help me please....I have geforce mx110


I want to do a deepfake desperately......please someone help me fix this.....
 

Thedarklook

DF Vagrant
jespo1352 said:
Can't seem to pull faces from videos, image sets, or images pulled from videos. It gets everything, just doesn't get faces. Tried different videos, sizes, names. I have Cuda 9.0, cuDNN 7.05 (with files copied to Cuda directory, GTX 750 Nvidia, win 10, and Microsoft visual 2015. 

Any assistance would be appreciated

I Get this info below along with sample image set 


C:\OpenFaceSwap\python\python-3.6.3.amd64\lib\site-packages\h5py\__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.
  from ._conv import register_converters as _register_converters
Using TensorFlow backend.
Input Directory: C:\OpenFaceSwap\imgA
Output Directory: C:\OpenFaceSwap\alignA
Filter: filter.jpg
Using json serializer
Starting, this may take a while...
Loading Extract from Extract_Align plugin...
100%|████████████████████████████████████████████████████████████████████████████████| 215/215 [01:33<00:00,  2.29it/s]
Alignments filepath: C:\OpenFaceSwap\imgA\alignments.json
Writing alignments to: C:\OpenFaceSwap\imgA\alignments.json
-------------------------
Images found:        215
Faces detected:      0
-------------------------
Done!
Press any key to continue . . .

very nice mate
 
Top