CoDeGAN：用于生成对抗网络的对比解耦
CoDeGAN: Contrastive Disentanglement for Generative Adversarial Network
AI总结：
针对GAN无监督解耦中的模式崩溃问题，提出CoDeGAN，通过将解耦约束从图像域放宽至特征域并集成自监督预训练，提升了训练稳定性与解耦能力，在多个基准上优于现有方法。
摘要：
解耦是可解释机器学习中的一个关键问题，也引起了计算机视觉界的广泛关注。许多现有的基于GAN的类别解耦（无监督）方法，如InfoGAN及其变体，主要旨在最大化生成图像与其潜在编码之间的互信息（MI）。然而，这种关注可能导致网络在给定相同潜在因子时倾向于生成高度相似的图像，潜在地引发模式崩溃。为缓解此问题，我们提出了CoDeGAN（用于生成对抗网络的对比解耦），在此我们将解耦约束从图像域放宽到特征域。这一修改不仅增强了GAN训练的稳定性，还提升了解耦能力。此外，我们将自监督预训练集成到CoDeGAN中以学习语义表征，显著促进无监督解耦。大量实验结果证明，我们的方法在多个基准上优于最先进的方法。代码可在https://github.com/learninginvision/CoDeGAN获取。
英文摘要：
Disentanglement, a critical concern in interpretable machine learning, has also garnered significant attention from the computer vision community. Many existing GAN-based class disentanglement (unsupervised) approaches, such as InfoGAN and its variants, primarily aim to maximize the mutual information (MI) between the generated image and its latent codes. However, this focus may lead to a tendency for the network to generate highly similar images when presented with the same latent class factor, potentially resulting in mode collapse or mode dropping. To alleviate this problem, we propose \texttt{CoDeGAN} (Contrastive Disentanglement for Generative Adversarial Networks), where we relax similarity constraints for disentanglement from the image domain to the feature domain. This modification not only enhances the stability of GAN training but also improves their disentangling capabilities. Moreover, we integrate self-supervised pre-training into CoDeGAN to learn semantic representations, significantly facilitating unsupervised disentanglement. Extensive experimental results demonstrate the superiority of our method over state-of-the-art approaches across multiple benchmarks. The code is available at https://github.com/learninginvision/CoDeGAN.
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。
