Overview
This course on Generative Adversarial Networks for Image Synthesis and Translation aims to teach students how to identify and explain the essential components of GANs, modify existing GAN implementations, design GANs for novel applications, and understand recent improvements in GAN loss functions. The course is designed for Data Scientists, researchers, and software developers familiar with keras, tensorflow, or similar recent Deep Learning tools. The teaching method involves exploring recent GAN progress through a model that generates faces conditional on desired features.
Syllabus
Introduction
Image Translation
Training Images
Conditional Gans
Supervisor vs Unsupervised
Unimodal vs multimodal
Multiple methods
Pigs to Pigs HD
Semantic Label Map
Training Methods
Examples
Shared latent space assumption
Constraints
Weights
Results
Unsupervised Multimodal Image Translation
Conclusion
Taught by
Open Data Science