Thin-Plate Spline Motion Model for Image Animation Deepfake | 2023
Learn how to install and use the Thin Plate Spline Motion Model for image animation Python application on Windows. With this deepfake technique, you can animate any image using another video as a reference. We will be installing this application on Windows 11, but the same process should work on Windows 10 as well.
To begin, download and install Anaconda and create a Python environment with version 3.9. Next, download the GitHub files and remove “torch” and “torchvision” from the requirements.txt file. Install all the remaining dependencies. Use the provided code to install “torch” and “torchvision”. Afterward, download the pretrained models from the link below and add them to the checkpoints folder.
Now, you are all set to proceed. Add your own images and videos to the assets folder and use the command below (with or without Nvidia graphics) to animate your image as a video. Replace the image and video name accordingly. Also please note that this process will take quite a some time.
Try it out and let us know your feedback.
- Install Torch
pip install torch==1.12.0+cu113 -f https://download.pytorch.org/whl/torch_stable.html
- Install Torchvision
pip install torchvision==0.13.0+cu113 -f https://download.pytorch.org/whl/torch_stable.html
- Run the demo using Nivida graphics card
python demo.py –config config/vox-256.yaml –checkpoint checkpoints/vox.pth.tar –source_image assets/image.jpg –driving_video assets/driving.mp4 –cpu
- Run the demo using CPU
python demo.py –config config/vox-256.yaml –checkpoint checkpoints/vox.pth.tar –source_image assets/image.jpg –driving_video assets/driving.mp4