【VQ-VAE-2论文精读】Generating Diverse High-Fidelity Images with VQ-VAE-2

【VQ-VAE-2论文精读】Generating Diverse High-Fidelity Images with VQ-VAE-2

• 0、前言
• Abstract
• 1 Introduction
• 2 Background
• • 2.1 Vector Quantized Variational AutoEncoder
• 3 Method
• • 3.1 Stage 1: Learning Hierarchical Latent Codes
  • 3.2 Stage 2: Learning Priors over Latent Codes
  • 3.3 Trading off Diversity with Classifier Based Rejection Sampling
• 4 Related Works
• 5 Experiments<