IB-GAN：基于信息瓶颈生成对抗网络的可解耦表示学习
IB-GAN: Disentangled Representation Learning with Information Bottleneck Generative Adversarial Networks
AI总结：
本文提出IB-GAN，利用信息瓶颈框架约束生成器中间层互信息，实现可解耦表示学习，在解耦得分上媲美β-VAE并优于InfoGAN，生成质量更佳。
摘要：
我们提出了一种新的基于GAN的无监督模型，用于可解耦表示学习。该新模型是在尝试将信息瓶颈（IB）框架应用于GAN优化过程中发现的，因此命名为IB-GAN。IB-GAN的架构与InfoGAN部分相似，但有一个关键区别：利用生成器的中间层来约束输入与生成输出之间的互信息。该中间随机层可充当可学习的潜在分布，与生成器以端到端的方式联合训练。因此，IB-GAN的生成器能够以可解耦且可解释的方式利用潜在空间。通过在dSprites和Color-dSprites数据集上的实验，我们证明IB-GAN在解耦得分上达到了与最先进的β-VAE相当的水平，并优于InfoGAN。此外，在CelebA和3D Chairs数据集上，就FID得分而言，IB-GAN生成的样本的视觉质量和多样性通常优于β-VAE和InfoGAN。
英文摘要：
We propose a new GAN-based unsupervised model for disentangled representation learning. The new model is discovered in an attempt to utilize the Information Bottleneck (IB) framework to the optimization of GAN, thereby named IB-GAN. The architecture of IB-GAN is partially similar to that of InfoGAN but has a critical difference; an intermediate layer of the generator is leveraged to constrain the mutual information between the input and the generated output. The intermediate stochastic layer can serve as a learnable latent distribution that is trained with the generator jointly in an end-to-end fashion. As a result, the generator of IB-GAN can harness the latent space in a disentangled and interpretable manner. With the experiments on dSprites and Color-dSprites dataset, we demonstrate that IB-GAN achieves competitive disentanglement scores to those of state-of-the-art \b{eta}-VAEs and outperforms InfoGAN. Moreover, the visual quality and the diversity of samples generated by IB-GAN are often better than those by \b{eta}-VAEs and Info-GAN in terms of FID score on CelebA and 3D Chairs dataset.
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。
