Deep Generative models (and specifically GANs) have seen amazing progress over the past few years. We are now at a point where these models can generate hyperrealistic images by training on image datasets without supervision.
But what do we do with these models? It seems like you can only use them to sample random images, right? Well, not entirely. It turns out that Deep Generative models learn a surprising amount of structure about the dataset they are trained on.
In this hack session I will explain how you can leverage this structure to deliberately manipulate image attributes by adjusting image representations in the latent space of a GAN. This hack session will use GPU-powered Google Colab notebooks so you can reproduce all the results for yourself!
Key Takeaways for the Audience
Powered by recent progressive in Deep Generative models, a new industry of generative media is emerging where images, video and even audio can be generated with specific purposes in mind. This hack session will show you:
- Basics of leveraging GANs for all kinds of content generation
- Outlook of the future applications we can expect
Check out the below video to know more about the session.