MLJun 6, 2017
GAN and VAE from an Optimal Transport Point of View
arXiv:1706.01807v163 citations
Originality Synthesis-oriented
AI Analysis
This work provides theoretical insights for researchers in generative modeling, but it is incremental as it revisits and simplifies existing ideas.
The paper revisits connections between Variational Autoencoders (VAE), Generative Adversarial Networks (GAN), and Minimum Kantorovitch Estimators (MKE) from an optimal transport perspective, aiming to clarify theoretical relationships in a simplified setup.
This short article revisits some of the ideas introduced in arXiv:1701.07875 and arXiv:1705.07642 in a simple setup. This sheds some lights on the connexions between Variational Autoencoders (VAE), Generative Adversarial Networks (GAN) and Minimum Kantorovitch Estimators (MKE).