In this issue:

• we explain Variational Autoencoders; 

• we discuss VQ-VAE, DeepMind’s variational autoencoder for large-scale image generation; 

• we explore Pixyz, a simple library for building generative models in PyTorch. 

All we need is love and be on top of ML knowledge, right? Happy pre-Valentine! We 🫀 and 🧠 you:

USE THIS 30% OFF TO SUBSCRIBE

💡 ML Concept of the Day: Understanding Variational Autoencoders  

In this issue of our generative models series, we would like to explore variational autoencoders (VAEs), which have become one of the most popular techniques in this area of deep learning. VAEs are incredibly elegant in their design and one of the simplest types of generative models that we can build →learn more about VAEs

🔎 ML Research You Should Know: DeepMind VQ-VAE is a Variational Autoencoder for Large Scale Image Generation   

DeepMind’s paper showed that the right VAE architecture could generate high fidelity images that rival more sophisticated models like generative adversarial networks (GANs) →diving deeper (Subscribe for only $35/year) 

🤖 ML Technology to Follow: Pixyz is a Simple Library for Building Generative Models in PyTorch    

Despite the recent popularity of generative models, their implementation continues to be relatively challenging. Recently, researchers from the University of Tokyo open-sourced Pixyz, a PyTorch-based framework for advancing research and implementation of generative models →what is the Pixyz’s architecture of Pixyz and how to use it

TheSequence is a summary of groundbreaking ML research papers, engaging explanations of ML concepts, and exploration of new ML frameworks and platforms. We keep you up-to-date with the main AI news, trends, and technology developments.