Generative models experiments for generating a new Miguel (joabreu).
poetry install- VAE with DNN-based Encoder/Decoder:
src.models.vae.LinearModel - VAE with CNN-based Encoder/Decoder:
src.models.vae.CNNModel
All experiments are available in Jupyter Notebook difmi.ipynb.
A single image was used for training, image augmentation using torchvision was performed to enhance results.
