Optimal Transport Based Generative Autoencoders
This addresses the problem of unstable generative modeling for researchers and practitioners by offering a more stable alternative to GANs, though it appears incremental as it builds on prior autoencoder and optimal transport methods.
The authors tackled the instability and convergence issues in GAN training by proposing two generative autoencoders, AE-OTtrans and AE-OTgen, based on optimal transport, which achieved higher quality and diversity images, surpassing GANs on MNIST and FashionMNIST datasets and setting state-of-the-art results on MNIST, FashionMNIST, and CelebA compared to non-adversarial models.
The field of deep generative modeling is dominated by generative adversarial networks (GANs). However, the training of GANs often lacks stability, fails to converge, and suffers from model collapse. It takes an assortment of tricks to solve these problems, which may be difficult to understand for those seeking to apply generative modeling. Instead, we propose two novel generative autoencoders, AE-OTtrans and AE-OTgen, which rely on optimal transport instead of adversarial training. AE-OTtrans and AEOTgen, unlike VAE and WAE, preserve the manifold of the data; they do not force the latent distribution to match a normal distribution, resulting in greater quality images. AEOTtrans and AE-OTgen also produce images of higher diversity compared to their predecessor, AE-OT. We show that AE-OTtrans and AE-OTgen surpass GANs in the MNIST and FashionMNIST datasets. Furthermore, We show that AE-OTtrans and AE-OTgen do state of the art on the MNIST, FashionMNIST, and CelebA image sets comapred to other non-adversarial generative models.