Augmented CycleGAN: Learning Many-to-Many Mappings from Unpaired Data

Amjad Almahairi, Sai Rajeshwar, Alessandro Sordoni, Philip Bachman, Aaron Courville
Proceedings of the 35th International Conference on Machine Learning, PMLR 80:195-204, 2018.

Abstract

Learning inter-domain mappings from unpaired data can improve performance in structured prediction tasks, such as image segmentation, by reducing the need for paired data. CycleGAN was recently proposed for this problem, but critically assumes the underlying inter-domain mapping is approximately deterministic and one-to-one. This assumption renders the model ineffective for tasks requiring flexible, many-to-many mappings. We propose a new model, called Augmented CycleGAN, which learns many-to-many mappings between domains. We examine Augmented CycleGAN qualitatively and quantitatively on several image datasets.

Cite this Paper

BibTeX
@InProceedings{pmlr-v80-almahairi18a, title = {Augmented {C}ycle{GAN}: Learning Many-to-Many Mappings from Unpaired Data}, author = {Almahairi, Amjad and Rajeshwar, Sai and Sordoni, Alessandro and Bachman, Philip and Courville, Aaron}, booktitle = {Proceedings of the 35th International Conference on Machine Learning}, pages = {195--204}, year = {2018}, editor = {Dy, Jennifer and Krause, Andreas}, volume = {80}, series = {Proceedings of Machine Learning Research}, month = {10--15 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v80/almahairi18a/almahairi18a.pdf}, url = {https://proceedings.mlr.press/v80/almahairi18a.html}, abstract = {Learning inter-domain mappings from unpaired data can improve performance in structured prediction tasks, such as image segmentation, by reducing the need for paired data. CycleGAN was recently proposed for this problem, but critically assumes the underlying inter-domain mapping is approximately deterministic and one-to-one. This assumption renders the model ineffective for tasks requiring flexible, many-to-many mappings. We propose a new model, called Augmented CycleGAN, which learns many-to-many mappings between domains. We examine Augmented CycleGAN qualitatively and quantitatively on several image datasets.} }
Endnote
%0 Conference Paper %T Augmented CycleGAN: Learning Many-to-Many Mappings from Unpaired Data %A Amjad Almahairi %A Sai Rajeshwar %A Alessandro Sordoni %A Philip Bachman %A Aaron Courville %B Proceedings of the 35th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2018 %E Jennifer Dy %E Andreas Krause %F pmlr-v80-almahairi18a %I PMLR %P 195--204 %U https://proceedings.mlr.press/v80/almahairi18a.html %V 80 %X Learning inter-domain mappings from unpaired data can improve performance in structured prediction tasks, such as image segmentation, by reducing the need for paired data. CycleGAN was recently proposed for this problem, but critically assumes the underlying inter-domain mapping is approximately deterministic and one-to-one. This assumption renders the model ineffective for tasks requiring flexible, many-to-many mappings. We propose a new model, called Augmented CycleGAN, which learns many-to-many mappings between domains. We examine Augmented CycleGAN qualitatively and quantitatively on several image datasets.
APA
Almahairi, A., Rajeshwar, S., Sordoni, A., Bachman, P. & Courville, A.. (2018). Augmented CycleGAN: Learning Many-to-Many Mappings from Unpaired Data. Proceedings of the 35th International Conference on Machine Learning, in Proceedings of Machine Learning Research 80:195-204 Available from https://proceedings.mlr.press/v80/almahairi18a.html.

Related Material