AlignFlow: Cycle Consistent Learning from Multiple Domains via Normalizing Flows

Abstract

Given datasets from multiple domains, a key challenge is to efficiently exploit these data sources for modeling a target domain. Variants of this problem have been studied in many contexts, such as cross-domain translation and domain adaptation. We propose AlignFlow, a generative modeling framework that models each domain via a normalizing flow. The use of normalizing flows allows for a) flexibility in specifying learning objectives via adversarial training, maximum likelihood estimation, or a hybrid of the two methods; and b) learning and exact inference of a shared representation in the latent space of the generative model. We derive a uniform set of conditions under which AlignFlow is marginally-consistent for the different learning objectives. Furthermore, we show that AlignFlow guarantees exact cycle consistency in mapping datapoints from a source domain to target and back to the source domain. Empirically, AlignFlow outperforms relevant baselines on image-to-image translation and unsupervised domain adaptation and can be used to simultaneously interpolate across the various domains using the learned representation.

Spectral Synthesis for Geostationary Satellite-to-Satellite Translation

“…However, there is a concern that this reduces data precision and fails to detect rare and anomalous events which may affect scientific applications. Extending our work to use normalizing flows, as in [16], may reduce this limitation.…”

“…We treat satellites with either dissimilar spectral coverage or varying vantage points as separate spectral sets. In this way, the problem closely resembles that of colorization [13] and imageto-image translation tasks [14]- [16] in the case where paired images are not available but with the added complexity of a large number of spectral bands. We use a combination of variational autoencoder (VAE) and generative adversarial network (GAN) [17] architectures adapted to our problem to model a shared latent space, as in unsupervised imageto-image translation [14].…”

Constrained Density Matching and Modeling for Cross-lingual Alignment of Contextualized Representations

“…From this, one can infer data likelihood of an unknown complex distribution using a simple Gaussian distribution. This technique has been shown useful in a range of cross-domain applications, such as image-to-image translation (Grover et al, 2020;Gong et al, 2019) and machine translation (Setiawan et al, 2020).…”

PairFlow: Enhancing Portable Chest X-Ray By Flow-Based Deformation For Covid-19 Diagnosing

“…The invertible mapping provides a distributional estimation of features in the latent space. Recently, many efforts making use of flow-based generative networks have been proposed to transfer between two unpaired data [10,11,12,13,14]. By using invertible properties, the flow-based methods can ensure exact cycle consistency in translation from a source domain to the target and returning to the source domain without any further loss functions.…”