arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2103.14017cs.CVcs.LG
收藏

图像平移的尺度放大解纠缠

Scaling-up Disentanglement for Image Translation

The Hebrew University of Jerusalem(耶路撒冷希伯来大学)

Aviv Gabbay, Yedid Hoshen

摘要:

图像翻译方法通常旨在操作一组标记属性(在训练时作为监督,例如域标签),同时保持未标记属性不变。当前的方法实现了以下任一种:(i)解缠结,其表现出低视觉保真度,并且仅在属性完全不相关的情况下才能满足。(ii)视觉上看似合理的翻译,显然没有解开。在这项工作中,我们提出了OverLORD,一个单一的框架,用于解开标记和未标记的属性以及合成高保真图像,它由两个阶段组成;(i)解开:学习解开的表示与潜在的优化。与以前的方法不同,我们不依赖于对抗性训练或任何架构偏见。(ii)合成:训练前馈编码器,用于推断学习的属性,并以对抗的方式调整生成器,以提高感知质量。当标记和未标记的属性是相关的,我们建立了一个额外的表示,占相关的属性,并提高解纠缠。我们强调,我们灵活的框架涵盖了多种设置,如解开标记的属性,姿势和外观,本地化的概念,形状和纹理。我们提出了显着更好的解纠缠更高的翻译质量和更大的输出多样性比国家的最先进的方法。

英文摘要:

Image translation methods typically aim to manipulate a set of labeled attributes (given as supervision at training time e.g. domain label) while leaving the unlabeled attributes intact. Current methods achieve either: (i) disentanglement, which exhibits low visual fidelity and can only be satisfied where the attributes are perfectly uncorrelated. (ii) visually-plausible translations, which are clearly not disentangled. In this work, we propose OverLORD, a single framework for disentangling labeled and unlabeled attributes as well as synthesizing high-fidelity images, which is composed of two stages; (i) Disentanglement: Learning disentangled representations with latent optimization. Differently from previous approaches, we do not rely on adversarial training or any architectural biases. (ii) Synthesis: Training feed-forward encoders for inferring the learned attributes and tuning the generator in an adversarial manner to increase the perceptual quality. When the labeled and unlabeled attributes are correlated, we model an additional representation that accounts for the correlated attributes and improves disentanglement. We highlight that our flexible framework covers multiple settings as disentangling labeled attributes, pose and appearance, localized concepts, and shape and texture. We present significantly better disentanglement with higher translation quality and greater output diversity than state-of-the-art methods.

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。

↑