arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2509.22169cs.CVcs.LG

DragGANSpace:面向 GAN 的潜空间探索与控制

DragGANSpace: Latent Space Exploration and Control for GANs

Georgia Institute of Technology(佐治亚理工学院)

Kirsten Odendaal, Neela Kaushik, Spencer Halverson

AI总结:

本文融合 StyleGAN、DragGAN 与 PCA,提出 DragGANSpace,通过潜空间降维与跨模型对齐提升 GAN 图像编辑的优化效率、视觉质量及可解释控制能力。

摘要:

本工作整合了 StyleGAN、DragGAN 和主成分分析(PCA),以提升 GAN 生成图像的潜空间效率与可控性。StyleGAN 提供结构化的潜空间,DragGAN 支持直观的图像操控,而 PCA 可降低维度并促进跨模型对齐,从而对潜空间进行更精简、更具可解释性的探索。我们将这些技术应用于 Animal Faces High Quality(AFHQ)数据集,发现将基于 PCA 的降维与用于图像操控的 DragGAN 框架相结合的方法,在保持性能的同时提高了优化效率。值得注意的是,将 PCA 引入 DragGAN 的潜 W+ 层能够持续减少总优化时间,同时保持良好的视觉质量,甚至提升优化后图像的结构相似性指数度量(SSIM),尤其是在较浅的潜空间中(W+ 层数 = 3)。我们还展示了对由两个在相似但不同的数据域(AFHQ-Dog 和 AFHQ-Cat)上训练的 StyleGAN 模型所生成图像进行对齐的能力,并表明我们能够控制这些对齐图像的潜空间,以直观且可解释的方式操控图像。我们的发现凸显了在广泛的图像合成与编辑应用中实现高效、可解释的潜空间控制的可能性。

英文摘要:

This work integrates StyleGAN, DragGAN and Principal Component Analysis (PCA) to enhance the latent space efficiency and controllability of GAN-generated images. Style-GAN provides a structured latent space, DragGAN enables intuitive image manipulation, and PCA reduces dimensionality and facilitates cross-model alignment for more streamlined and interpretable exploration of latent spaces. We apply our techniques to the Animal Faces High Quality (AFHQ) dataset, and find that our approach of integrating PCA-based dimensionality reduction with the Drag-GAN framework for image manipulation retains performance while improving optimization efficiency. Notably, introducing PCA into the latent W+ layers of DragGAN can consistently reduce the total optimization time while maintaining good visual quality and even boosting the Structural Similarity Index Measure (SSIM) of the optimized image, particularly in shallower latent spaces (W+ layers = 3). We also demonstrate capability for aligning images generated by two StyleGAN models trained on similar but distinct data domains (AFHQ-Dog and AFHQ-Cat), and show that we can control the latent space of these aligned images to manipulate the images in an intuitive and interpretable manner. Our findings highlight the possibility for efficient and interpretable latent space control for a wide range of image synthesis and editing applications.

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。

↑