SYS863C - Sujets spéciaux I en G.P.A. : Deep Generative Modeling: Theory and Applications
Deep generative models have emerged rapidly as the state-of-the-art techniques for generating real-looking image and non-image data samples. Their applications comprise image editing, data augmentation (e.g., for enabling tumor classification in case of low data schemes in medicine), image translation, preserving privacy, and so forth.
This course aims to familiarize students with different deep generative models including (but not limited to) Generative Adversarial Network (GAN) and Variational Autoencoder (VAE). In the first part of the course, the students would learn the basics and theory behind these models.
In the second part of the course, the applications of generative models would be discussed. At the end of this course, the student will be able to:
(i) understand the strengths and limitations of different types of deep generative models;
(ii) customize and apply these models to their research works.