# SPA-GAN: Spatial Attention GAN for Image-to-Image Translation

## TL;DR

SPA-GAN introduces a spatial attention mechanism within the discriminator to enhance image-to-image translation, focusing on discriminative regions and preserving domain features, resulting in superior, lightweight translation performance.

## Contribution

The paper presents a novel SPA-GAN model that integrates attention directly into the discriminator without extra networks, improving translation quality and efficiency.

## Key findings

- SPA-GAN outperforms state-of-the-art methods on benchmark datasets.
- It is a lightweight model without additional attention networks.
- Incorporates feature map loss to preserve domain-specific features.

## Abstract

Image-to-image translation is to learn a mapping between images from a source domain and images from a target domain. In this paper, we introduce the attention mechanism directly to the generative adversarial network (GAN) architecture and propose a novel spatial attention GAN model (SPA-GAN) for image-to-image translation tasks. SPA-GAN computes the attention in its discriminator and use it to help the generator focus more on the most discriminative regions between the source and target domains, leading to more realistic output images. We also find it helpful to introduce an additional feature map loss in SPA-GAN training to preserve domain specific features during translation. Compared with existing attention-guided GAN models, SPA-GAN is a lightweight model that does not need additional attention networks or supervision. Qualitative and quantitative comparison against state-of-the-art methods on benchmark datasets demonstrates the superior performance of SPA-GAN.