当前位置:
X-MOL 学术
›
arXiv.cs.CV
›
论文详情
Our official English website, www.x-mol.net, welcomes your
feedback! (Note: you will need to create a separate account there.)
Guided Disentanglement in Generative Networks
arXiv - CS - Computer Vision and Pattern Recognition Pub Date : 2021-07-29 , DOI: arxiv-2107.14229 Fabio Pizzati, Pietro Cerri, Raoul de Charette
arXiv - CS - Computer Vision and Pattern Recognition Pub Date : 2021-07-29 , DOI: arxiv-2107.14229 Fabio Pizzati, Pietro Cerri, Raoul de Charette
Image-to-image translation (i2i) networks suffer from entanglement effects in
presence of physics-related phenomena in target domain (such as occlusions,
fog, etc), thus lowering the translation quality and variability. In this
paper, we present a comprehensive method for disentangling physics-based traits
in the translation, guiding the learning process with neural or physical
models. For the latter, we integrate adversarial estimation and genetic
algorithms to correctly achieve disentanglement. The results show our approach
dramatically increase performances in many challenging scenarios for image
translation.
中文翻译:
生成网络中的引导解开
图像到图像转换 (i2i) 网络在目标域中存在与物理相关的现象(例如遮挡、雾等)时会受到纠缠效应的影响,从而降低了转换质量和可变性。在本文中,我们提出了一种在翻译中解开基于物理的特征的综合方法,用神经或物理模型指导学习过程。对于后者,我们整合了对抗性估计和遗传算法以正确实现解缠结。结果表明,我们的方法在许多具有挑战性的图像翻译场景中显着提高了性能。
更新日期:2021-07-30
中文翻译:
生成网络中的引导解开
图像到图像转换 (i2i) 网络在目标域中存在与物理相关的现象(例如遮挡、雾等)时会受到纠缠效应的影响,从而降低了转换质量和可变性。在本文中,我们提出了一种在翻译中解开基于物理的特征的综合方法,用神经或物理模型指导学习过程。对于后者,我们整合了对抗性估计和遗传算法以正确实现解缠结。结果表明,我们的方法在许多具有挑战性的图像翻译场景中显着提高了性能。