The CelebA dataset (https://www.kaggle.com/datasets/jessicali9530/celeba-dataset) was used to generate images with two different approaches, using Variational Autoencoders (VAEs) and using Generative Adversarial Networks (GANs). This was just an exercise so results are not good. For to achieve good results, much more iterations and time working on the architectures are needed.
-
Notifications
You must be signed in to change notification settings - Fork 0
kirle/CelebA_FaceGeneration
About
Using VAEs and GANs to understand how to generate images, over the CelebA dataset
Topics
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published