Argmax Flows and Multinomial Diffusion: Learning Categorical DistributionsDownload PDF

21 May 2021, 20:41 (edited 22 Oct 2021)NeurIPS 2021 PosterReaders: Everyone
  • Keywords: categorical, normalizing flows, diffusion
  • TL;DR: Extensions of normalizing flows and diffusion for categorical data
  • Abstract: Generative flows and diffusion models have been predominantly trained on ordinal data, for example natural images. This paper introduces two extensions of flows and diffusion for categorical data such as language or image segmentation: Argmax Flows and Multinomial Diffusion. Argmax Flows are defined by a composition of a continuous distribution (such as a normalizing flow), and an argmax function. To optimize this model, we learn a probabilistic inverse for the argmax that lifts the categorical data to a continuous space. Multinomial Diffusion gradually adds categorical noise in a diffusion process, for which the generative denoising process is learned. We demonstrate that our method outperforms existing dequantization approaches on text modelling and modelling on image segmentation maps in log-likelihood.
  • Supplementary Material: pdf
  • Code Of Conduct: I certify that all co-authors of this work have read and commit to adhering to the NeurIPS Statement on Ethics, Fairness, Inclusivity, and Code of Conduct.
  • Code:
12 Replies