Work

ALICE: Towards Understanding Adversarial Learning for Joint Distribution Matching

Year: 2017

Type: preprint

Abstract: We investigate the non-identifiability issues associated with bidirectional adversarial training for joint distribution matching. Within a framework of conditional entropy, we propose both adversarial and non-adversarial approaches to learn desirable matched joint distributions for unsupervised and supervised tasks. We unify a broad family of adversarial models as joint distribution matching problems. Our approach stabilizes learning of unsupervised bidirectional adversarial learning methods. Further, we introduce an extension for semi-supervised learning tasks. Theoretical results are validated in synthetic data and real-world applications.

Source: arXiv (Cornell University)

DOI: 10.48550/arxiv.1709.01215

Authors: Li, Chunyuan, Liu, Hao, Chen, Changyou, Pu, Yunchen, Chen, Liqun, Henao, Ricardo, Carin, Lawrence

Institutions: Duke University, Nanjing University, University at Buffalo, State University of New York, University of Surrey

Language: English

Topic: Generative Adversarial Networks and Image Synthesis

Subfield: Computer Vision and Pattern Recognition

Field: Computer Science

Domain: Physical Sciences

Open Access status: green

Funders: National Science Foundation, U.S. Department of Energy, Defense Advanced Research Projects Agency, Office of Naval Research, Army Research Office

OpenAlex ID: W2963592272

Cites:

Cited by: 190

Related to: 10