Drag Your GAN: Interactive Point-based Manipulation on the Generative Image Manifold
SIGGRAPH 2023 Conference Proceedings
Web Demos
Requirements
If you have CUDA graphic card, please follow the requirements of NVlabs/stylegan3.
The usual installation steps involve the following commands, they should set up the correct CUDA version and all the python packages
Then install the additional requirements
Otherwise (for GPU acceleration on MacOS with Silicon Mac M1/M2, or just CPU) try the following:
Run Gradio visualizer in Docker
Provided docker image is based on NGC PyTorch repository. To quickly try out visualizer in Docker, run the following:
Now you can open a shared link from Gradio (printed in the terminal console).
Beware the Docker image takes about 25GB of disk space!
Download pre-trained StyleGAN2 weights
To download pre-trained weights, simply run:
If you want to try StyleGAN-Human and the Landscapes HQ (LHQ) dataset, please download weights from these links: StyleGAN-Human, LHQ, and put them under
./checkpoints.Feel free to try other pretrained StyleGAN.
Run DragGAN GUI
To start the DragGAN GUI, simply run:
If you are using windows, you can run:
This GUI supports editing GAN-generated images. To edit a real image, you need to first perform GAN inversion using tools like PTI. Then load the new latent code and model weights to the GUI.
You can run DragGAN Gradio demo as well, this is universal for both windows and linux:
Acknowledgement
This code is developed based on StyleGAN3. Part of the code is borrowed from StyleGAN-Human.
(cheers to the community as well)
License
The code related to the DragGAN algorithm is licensed under CC-BY-NC. However, most of this project are available under a separate license terms: all codes used or modified from StyleGAN3 is under the Nvidia Source Code License.
Any form of use and derivative of this code must preserve the watermarking functionality showing "AI Generated".