An Optimized Architecture for Unpaired Image-to-Image Translation
This work addresses optimization in training for image translation tasks, but it appears incremental as it builds on Cycle-GAN with specific improvements.
The paper tackled the problem of unpaired image-to-image translation by proposing a new neural network architecture that eliminates the need for reverse mapping, resulting in significantly reduced training duration.
Unpaired Image-to-Image translation aims to convert the image from one domain (input domain A) to another domain (target domain B), without providing paired examples for the training. The state-of-the-art, Cycle-GAN demonstrated the power of Generative Adversarial Networks with Cycle-Consistency Loss. While its results are promising, there is scope for optimization in the training process. This paper introduces a new neural network architecture, which only learns the translation from domain A to B and eliminates the need for reverse mapping (B to A), by introducing a new Deviation-loss term. Furthermore, few other improvements to the Cycle-GAN are found and utilized in this new architecture, contributing to significantly lesser training duration.