SDA-GAN: Unsupervised Image Translation Using Spectral Domain Attention-Guided Generative Adversarial Network

Abstract

This work introduced a novel GAN architecture for unsupervised image translation on the task of face style transform. A spectral attention-based mechanism is embedded into the design along with spatial attention on the image contents. We proved that neural network has the potential of learning complex transformations such as Fourier transform, within considerable computational cost. The model is trained and tested in comparison to the baseline model, which only uses spatial attention. The performance improvement of our approach is significant especially when the source and target domain include different complexity (reduced FID to 49.18 from 142.84). In the translation process, a spectra filling effect was introduced due to the implementation of FFT and spectral attention. Another style transfer task and real-world object translation are also studied in this paper.

arXiv comment7 pages, 3 figures

References in corpus (9)

• GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
• Self-Attention Generative Adversarial Networks
• Generative Adversarial Text to Image Synthesis
• Augmented CycleGAN: Learning Many-to-Many Mappings from Unpaired Data
• On Convergence and Stability of GANs
• DRIT++: Diverse Image-to-Image Translation via Disentangled Representations
• Coupled Generative Adversarial Networks
• AttentionGAN: Unpaired Image-to-Image Translation using Attention-Guided Generative Adversarial Networks
• GMM-UNIT: Unsupervised Multi-Domain and Multi-Modal Image-to-Image Translation via Attribute Gaussian Mixture Modeling