Creating Endless Anime Faces with Python + StyleGAN2 + Colab

🐶What’s cuter than an anime girl? Infinite anime girls. Here we look at how to code an anime face generator using Python and a ready-trained anime data model. We’ll be doing it in Google Colab but you could run it anywhere. We also look at how to save the faces being generated and prototype a basic web app where a visitor can click a button and get a new face.

00:05 Intro
02:35 Setting up face generation
08:30 Creating a single face
12:45 Creating a grid of faces
17:02 Saving faces to Google Drive
22:00 Creating a web app
30:00 Exposing the web app with Ngrok
35:20 Summary

You can get the entire .ipynb file here: https://colab.research.google.com/dri...

A simple web-app showcasing faces: https://practical-meitner-69afb9.netl...

The training data is available both here: https://drive.google.com/open?id=1WNQ...
and here:
https://drive.google.com/file/d/1NI7z...

The first part draws heavily on the 2020 article by Fathy Rashad: https://towardsdatascience.com/genera...

Learn more about StyleGAN2 here:
https://github.com/NVlabs/stylegan2

Streamlit: https://streamlit.io/

Ngrok: https://ngrok.com/

Subscribe for more: @mikaelcodes

#Python #MachineLearning #Anime #Colab #TensorFlow #StyleGAN2 #FaceGeneration #AIArt #Ngrok #Waifu #Streamlit #WebApp