Deepfake(2): MakeItTalk-Python deepfake image talk

For Face-Morphing, clone using the github command in the following location.

https://github.com/Azmarie/Face-Morphing

Of course, use the command git clone https://github.com/Azmarie/Face-Morphing.

The required packages are:

numpy
scikit_image
opencv_python
Pillow
skimage
dlib (요건 Cmake 패키지를 먼저 설치하라고 함)

It will ask you to install “imutils” package later.

Even though it’s been mentioned previously, here are the steps to clone the source code from GitHub, create and activate a dedicated environment using Conda, and install the required libraries:

  • Clone source code from GitHub:
    Clone the code to your local system using the git clone https://github.com/Azmarie/Face-Morphing.git command in the terminal.
  • Create and enable Conda environment:
    Create a new environment with the command “conda create -n FaceMorph python=3.7.9” in the terminal. To activate the created environment, type “conda activate FaceMorph”
  • Install dependencies using txt:
    Go to the directory where your requirements.txt file is located.
    Install the necessary libraries using the “pip install -r requirements.txt” command.
  • Disable environment (optional):
    After completing the task, deactivate the environment with the c”onda deactivate” command.

After installation, additional installation of libraries was required, so the changed library result dependencies can be obtained by “pip freeze>requirements.txt”. We recommend that you save this content as requirements.txt and try installing it in your own environment. (if the original requirements.txt doesn’t work)

The result obtained by “pip freeze>requirements.txt” is shown below.

certifi==2020.6.20
cmake==3.22.3
dlib==19.23.1
imageio==2.16.1
imutils==0.5.4
networkx==2.6.3
numpy==1.21.5
opencv-python==4.5.5.64
packaging==21.3
Pillow==9.0.1
pyparsing==3.0.7
PyWavelets==1.3.0
scikit-image==0.19.2
scipy==1.7.3
tifffile==2021.11.2

python3 code/__init__.py --img1 images/aligned_images/jennie.png --img2 images/aligned_images/rih.png --output output.mp4

jennie.png and rih.png will be morphed and output to output.mp4

I commented on these three lines in face_morph.py. The image pixel numbers should be correspondent.

In Gimp

Leave a Reply

Your email address will not be published. Required fields are marked *