BabyGAN

StyleGAN-based predictor of children's faces from photographs of theoretical parents. The latent representation is extracted from the input images, after which the algorithm mixes them in certain proportions. The neural network model is based on the GAN architecture. Using latency direction, you can change various parameters: age, face position, emotions and gender.

example3

example1

example2

Pre-train Models and dictionaries

Follow the LINK and add shortcut to Drive:

mount_eng

The folder structure should be:

.
├── data                    
│   └── finetuned_resnet.h5 
├── karras2019stylegan-ffhq-1024x1024.pkl
├── shape_predictor_68_face_landmarks.dat.bz2
├── vgg16_weights_tf_dim_ordering_tf_kernels_notop.h5
├── vgg16_zhang_perceptual.pkl
└── ...

Prerequisites

  • 64-bit Python 3.6 installation.
  • TensorFlow 1.10.0 with GPU support.
  • One or more high-end NVIDIA GPUs with at least 11GB of DRAM.
  • NVIDIA driver 391.35 or newer, CUDA toolkit 9.0 or newer, cuDNN 7.3.1 or newer.

Generating latent representation of your images

You can generate latent representations of your own images using two scripts:

  1. Create folders for photos

mkdir raw_images aligned_images

  1. Extract and align faces from images

python align_images.py raw_images/ aligned_images/

  1. Find latent representation of aligned images

python encode_images.py aligned_images/ generated_images/ latent_representations/

GitHub