Estimating the Success of Unsupervised Image to Image Translation

Sagie Benaim, Tomer Galanti, Lior Wolf

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review


While in supervised learning, the validation error is an unbiased estimator of the generalization (test) error and complexity-based generalization bounds are abundant, no such bounds exist for learning a mapping in an unsupervised way. As a result, when training GANs and specifically when using GANs for learning to map between domains in a completely unsupervised way, one is forced to select the hyperparameters and the stopping epoch by subjectively examining multiple options. We propose a novel bound for predicting the success of unsupervised cross domain mapping methods, which is motivated by the recently proposed Simplicity Principle. The bound can be applied both in expectation, for comparing hyperparameters and for selecting a stopping criterion, or per sample, in order to predict the success of a specific cross-domain translation. The utility of the bound is demonstrated in an extensive set of experiments employing multiple recent algorithms. Our code is available at

Original languageEnglish
Title of host publicationComputer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings
EditorsVittorio Ferrari, Cristian Sminchisescu, Martial Hebert, Yair Weiss
Number of pages17
StatePublished - 2018
Event15th European Conference on Computer Vision, ECCV 2018 - Munich, Germany
Duration: 8 Sep 201814 Sep 2018

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11209 LNCS


Conference15th European Conference on Computer Vision, ECCV 2018


  • GANs
  • Generalization bounds
  • Image to image translation
  • Unsupervised learning

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)


Dive into the research topics of 'Estimating the Success of Unsupervised Image to Image Translation'. Together they form a unique fingerprint.

Cite this