CVJun 16, 2018

Show, Attend and Translate: Unsupervised Image Translation with Self-Regularization and Attention

arXiv:1806.06195v353 citations
Originality Incremental advance
AI Analysis

This addresses the problem of unwanted changes in image translation for applications like data augmentation, with incremental improvements in method design.

The paper tackles unsupervised image translation by proposing a model with self-regularization and an attention module to reduce artifacts, achieving significantly better performance than existing methods.

Image translation between two domains is a class of problems aiming to learn mapping from an input image in the source domain to an output image in the target domain. It has been applied to numerous domains, such as data augmentation, domain adaptation, and unsupervised training. When paired training data is not accessible, image translation becomes an ill-posed problem. We constrain the problem with the assumption that the translated image needs to be perceptually similar to the original image and also appears to be drawn from the new domain, and propose a simple yet effective image translation model consisting of a single generator trained with a self-regularization term and an adversarial term. We further notice that existing image translation techniques are agnostic to the subjects of interest and often introduce unwanted changes or artifacts to the input. Thus we propose to add an attention module to predict an attention map to guide the image translation process. The module learns to attend to key parts of the image while keeping everything else unaltered, essentially avoiding undesired artifacts or changes. The predicted attention map also opens door to applications such as unsupervised segmentation and saliency detection. Extensive experiments and evaluations show that our model while being simpler, achieves significantly better performance than existing image translation methods.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes