NoisyTwins：通过StyleGAN实现类一致且多样化的图像生成
NoisyTwins: Class-Consistent and Diverse Image Generation through StyleGANs
AI总结：
NoisyTwins针对StyleGAN在长尾数据集上类别条件训练时潜在变量坍缩的问题，提出类别嵌入增强与自监督潜在解相关方法，在ImageNet-LT和iNaturalist 2019上FID提升约19%，实现类一致且多样化的图像生成。
摘要：
StyleGAN处于可控图像生成的前沿，因为它们产生的潜在空间在语义上是解耦的，使其适用于图像编辑和操作。然而，当在大规模长尾数据集上通过类别条件进行训练时，StyleGAN的性能会严重下降。我们发现性能下降的一个原因是每个类别在$\mathcal{W}$潜在空间中的潜在变量坍缩。借助NoisyTwins，我们首先为类别嵌入引入一种有效且低成本的增强策略，然后基于$\mathcal{W}$空间中的自监督来解相关潜在变量。这种解相关缓解了坍缩问题，确保我们的方法在图像生成中保持类内多样性和类一致性。我们在ImageNet-LT和iNaturalist 2019这两个大规模真实世界长尾数据集上展示了我们方法的有效性，在这些数据集上，我们的方法在FID指标上比其他方法高出约19%，建立了新的最先进水平。
英文摘要：
StyleGANs are at the forefront of controllable image generation as they produce a latent space that is semantically disentangled, making it suitable for image editing and manipulation. However, the performance of StyleGANs severely degrades when trained via class-conditioning on large-scale long-tailed datasets. We find that one reason for degradation is the collapse of latents for each class in the $\mathcal{W}$ latent space. With NoisyTwins, we first introduce an effective and inexpensive augmentation strategy for class embeddings, which then decorrelates the latents based on self-supervision in the $\mathcal{W}$ space. This decorrelation mitigates collapse, ensuring that our method preserves intra-class diversity with class-consistency in image generation. We show the effectiveness of our approach on large-scale real-world long-tailed datasets of ImageNet-LT and iNaturalist 2019, where our method outperforms other methods by $\sim 19\%$ on FID, establishing a new state-of-the-art.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
