Abstract
Domain translation is the task of finding correspondence between two domains. Several deep neural network (DNN) models, e.g., CycleGAN and cross-lingual language models, have shown remarkable successes on this task under the unsupervised setting--the mappings between the domains are learned from two independent sets of training data in both domains (without paired samples). However, those methods typically do not perform well on a significant proportion of test samples. In this article, we hypothesize that many of such unsuccessful samples lie at the fringe--relatively low-density areas--of data distribution, where the DNN was not trained very well, and propose to perform the Langevin dynamics to bring such fringe samples toward high-density areas. We demonstrate qualitatively and quantitatively that our strategy, called Langevin cooling (L-Cool), enhances state-of-the-art methods in image translation and language translation tasks.
Highlights
R ECENTLY, deep neural networks (DNNs) have broadly contributed across various application domains in the Manuscript received August 13, 2020; revised March 22, 2021 and November 12, 2021; accepted January 11, 2022
Klaus-Robert Müller is with the Machine Learning Group, Technische Universität Berlin, 10587 Berlin, Germany, with the Berlin Institute for the Foundations of Learning and Data (BIFOLD), 10587 Berlin, Germany, with the Department of Artificial Intelligence, Korea University, Seoul 136-713, South Korea, and with the Max Planck Institute for Informatics, 66123 Saarbrücken, Germany (e-mail: klaus-robert.mueller@ tu-berlin.de)
Similar ideas were applied to natural language processing (NLP): dual learning [18], [35] and cross-lingual language models (XLMs) [19], which are trained on unpaired monolingual data and achieved high performance in language translation
Summary
R ECENTLY, deep neural networks (DNNs) have broadly contributed across various application domains in the Manuscript received August 13, 2020; revised March 22, 2021 and November 12, 2021; accepted January 11, 2022. CycleGAN, an extension of generative adversarial networks (GANs) [31], showed its capability of unsupervised DT with impressive results in image translation tasks [32]–[34] It learns the mappings between the two domains by matching the source training distribution transferred to the target domain and the target training distribution under the cycle-consistency constraint. Similar ideas were applied to natural language processing (NLP): dual learning [18], [35] and cross-lingual language models (XLMs) [19], which are trained on unpaired monolingual data and achieved high performance in language translation Despite their remarkable successes, existing unsupervised DT methods are known to fail on a significant proportion of test samples [32], [36]–[38]. 2) Qualitative evaluation of L-Cool in comparison with state-of-the-art methods as well as image processing techniques (as baseline projection methods) on image
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: IEEE transactions on neural networks and learning systems
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
