简化GAN架构，证明仅用对抗损失即可实现高效图像翻译。
Mechanisms of Generative Image-to-Image Translation Networks
- 采用更简单的网络结构，仅依赖GAN损失进行图像翻译。
- 在多个数据集上达到与复杂模型相当的生成效果。
- 适合希望快速部署、避免繁琐损失设计的研究者。
生成对抗网络（GAN）在图像到图像翻译任务中广泛应用。本文提出一种结构更简洁的图像翻译网络，相比现有模型更为轻量。我们研究了GAN与自编码器之间的关系，解释了为何仅使用GAN组件即可有效完成图像翻译任务。实验表明，仅通过对抗损失即可获得与复杂模型相当的结果，无需额外设计复杂的损失惩罚项。我们进一步阐明了该现象背后的原理，并通过实验证明了结论的有效性。
原文摘要 · Abstract (English)
Generative Adversarial Networks (GANs) are a class of neural networks that have been widely used in the field of image-to-image translation. In this paper, we propose a streamlined image-to-image translation network with a simpler architecture compared to existing models. We investigate the relationship between GANs and autoencoders and provide an explanation for the efficacy of employing only the GAN component for tasks involving image translation. We show that adversarial for GAN models yields results comparable to those of existing methods without additional complex loss penalties. Subsequently, we elucidate the rationale behind this phenomenon. We also incorporate experimental results to demonstrate the validity of our findings.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。