Keywords: adversarial, score matching, Langevin dynamics, GAN, generative model
Abstract: Denoising Score Matching with Annealed Langevin Sampling (DSM-ALS) has recently found success in generative modeling. The approach works by first training a neural network to estimate the score of a distribution, and then using Langevin dynamics to sample from the data distribution assumed by the score network. Despite the convincing visual quality of samples, this method appears to perform worse than Generative Adversarial Networks (GANs) under the Fréchet Inception Distance, a standard metric for generative models. We show that this apparent gap vanishes when denoising the final Langevin samples using the score network. In addition, we propose two improvements to DSM-ALS: 1) Consistent Annealed Sampling as a more stable alternative to Annealed Langevin Sampling, and 2) a hybrid training formulation, composed of both Denoising Score Matching and adversarial objectives. By combining these two techniques and exploring different network architectures, we elevate score matching methods and obtain results competitive with state-of-the-art image generation on CIFAR-10.
One-sentence Summary: Combining GANs with score matching and using Consistent Sampling (as an alternative to Langevin dynamics) for improved generative modeling
Supplementary Material: zip
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics
Code: [![github](/images/github_icon.svg) AlexiaJM/AdversarialConsistentScoreMatching](https://github.com/AlexiaJM/AdversarialConsistentScoreMatching)
Data: [CIFAR-10](https://paperswithcode.com/dataset/cifar-10), [MNIST](https://paperswithcode.com/dataset/mnist), [Stacked MNIST](https://paperswithcode.com/dataset/stacked-mnist)