Unpaired image-to-image translation using Generative Adversarial Networks (GAN) is successful in converting images among multiple domains. Moreover, recent studies have shown a way to diversify the outputs of the generator. However, since there are no restrictions on how the generator diversifies the results, it is likely to translate some unexpected features. In this paper, we propose Style-Restricted GAN (SRGAN) to demonstrate the importance of controlling the encoded features used in style diversifying process. More specifically, instead of KL divergence loss, we adopt three new losses to restrict the distribution of the encoded features: batch KL divergence loss, correlation loss, and histogram imitation loss. Further, the encoder is pre-trained with classification tasks before being used in translation process. The study reports quantitative as well as qualitative results with Precision, Recall, Density, and Coverage. The proposed three losses lead to the enhancement of the level of diversity compared to the conventional KL loss. In particular, SRGAN is found to be successful in translating with higher diversity and without changing the class-unrelated features in the CelebA face dataset. To conclude, the importance of the encoded features being well-regulated was proven with two experiments. Our implementation is available at https://github.com/shinshoji01/Style-Restricted_GAN.
翻译:使用 General Adversarial Networks (GAN) 将图像转换为图像, 使用 General Adversarial 网络( GAN) 将图像转换成图像, 取得了成功。 此外, 最近的研究展示了使生成器产出多样化的一种方法。 但是, 由于对生成器如何使结果多样化没有限制, 它可能会翻译出一些出人意料的特征。 在本文中, 我们提议Style- Restricted GAN( SRGAN) 以显示在风格多样化过程中所使用的编码特性的重要性。 更具体地说, 我们采用三种新的损失来限制编码特性的分布: 批量 KL 差异损失、 相关损失 和 直方图像模拟损失 。 此外, 在翻译过程中使用之前, 加密器对分类任务进行了预先训练。 研究报告报告了精度、 回调、 密度和覆盖范围的定量结果。 提议的三种损失导致多样性水平的提高, 而不是常规的 KLLL 差异损失。 特别是, SRGANAN 被发现成功地翻译了更高的多样性, 并且没有改变我们现有的 CelebA/ regrodrodrodududustration 的Card 。