Estimating the Success of Unsupervised Image to Image Translation
This addresses the problem of hyperparameter selection and stopping criteria in unsupervised GAN training for domain mapping, offering a practical tool for researchers and practitioners, though it is incremental as it builds on existing principles.
The paper tackles the lack of generalization bounds for unsupervised image-to-image translation, proposing a novel bound based on the Simplicity Principle to predict success, which is demonstrated through extensive experiments with multiple algorithms.
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 https://github.com/sagiebenaim/gan_bound .