Interactively Assessing Disentanglement in GANs
Interactively Assessing Disentanglement in GANs
复制标题
DOI：
10.1111/cgf.14524
复制
发表时间：
2022-06
影响因子：
2.5
通讯作者：
S. Jeong;Shusen Liu;Matthew Berger
中科院分区：
文献类型：
--
作者：
S. Jeong;Shusen Liu;Matthew Berger
生成对抗网络（GAN）近年来取得了巨大的发展，在许多领域显示出广泛的适用性。然而，对于人们来说，gan仍然是出了名的难以理解，特别是对于能够生成照片逼真图像的现代gan。在这项工作中，我们为GAN的可解释性提供了一种可视化分析方法，我们专注于GAN解纠缠的分析和可视化。解纠缠涉及到沿少量不同的语义变化因素控制GAN产生的内容的能力。我们的方法的目标是揭示GAN解纠缠的洞察力，超越粗糙的总结，而不是允许对GAN建模的数据分布进行更深入的分析。我们的可视化允许人们在数据分布的分组和趋势方面评估单个变化因素，其中我们的分析试图将gan的学习表示空间与gan产生的基于属性的图像语义评分联系起来。通过用例，我们表明我们的可视化在评估解缠方面是有效的，允许人们快速识别变异因素及其整体质量。此外，我们还展示了我们的方法如何突出gan学习到的潜在数据集偏差。
Generative adversarial networks (GAN) have witnessed tremendous growth in recent years, demonstrating wide applicability in many domains. However, GANs remain notoriously difficult for people to interpret, particularly for modern GANs capable of generating photo‐realistic imagery. In this work we contribute a visual analytics approach for GAN interpretability, where we focus on the analysis and visualization of GAN disentanglement. Disentanglement is concerned with the ability to control content produced by a GAN along a small number of distinct, yet semantic, factors of variation. The goal of our approach is to shed insight on GAN disentanglement, above and beyond coarse summaries, instead permitting a deeper analysis of the data distribution modeled by a GAN. Our visualization allows one to assess a single factor of variation in terms of groupings and trends in the data distribution, where our analysis seeks to relate the learned representation space of GANs with attribute‐based semantic scoring of images produced by GANs. Through use‐cases, we show that our visualization is effective in assessing disentanglement, allowing one to quickly recognize a factor of variation and its overall quality. In addition, we show how our approach can highlight potential dataset biases learned by GANs.