arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2103.15017cs.CVcs.AIcs.LG
收藏

H-GAN:GANs的力量掌握在你手中

H-GAN: the power of GANs in your Hands

University of Alicante(阿利坎特大学);FORTH(希腊研究与技术基金会);Institute of Space Sciences (ICE-CSIC)(空间科学研究所(ICE-CSIC))
Sergiu Oprea, Giorgos Karvounas, Pablo Martinez-Gonzalez, Nikolaos Kyriazis, Sergio Orts-Escolano, Iason Oikonomidis, Alberto Garcia-Garcia, Aggeliki Tsoli, Jos… 展开作者

Sergiu Oprea, Giorgos Karvounas, Pablo Martinez-Gonzalez, Nikolaos Kyriazis, Sergio Orts-Escolano, Iason Oikonomidis, Alberto Garcia-Garcia, Aggeliki Tsoli, Jose Garcia-Rodriguez, Antonis Argyros

摘要:

我们提出了HandGAN(H-GAN),这是一种实现多尺度感知鉴别器的周期一致性对抗学习方法。它的目的是将合成图像的手到真正的领域。合成手提供了完整的地面实况注释,但它们并不代表真实世界数据的目标分布。我们致力于提供逼真的手部外观与合成注释的完美融合。依靠图像到图像的翻译,我们改进了合成手的外观,以近似手的真实图像集合的统计分布。H-GAN不仅解决了跨域色调映射,还解决了局部区域中的结构差异,例如阴影不连续性。结果进行了评估,在定性和定量的基础上,提高以前的作品。此外,我们依赖于手部分类任务来声称我们生成的手部在统计上与手部的真实领域相似。

英文摘要:

We present HandGAN (H-GAN), a cycle-consistent adversarial learning approach implementing multi-scale perceptual discriminators. It is designed to translate synthetic images of hands to the real domain. Synthetic hands provide complete ground-truth annotations, yet they are not representative of the target distribution of real-world data. We strive to provide the perfect blend of a realistic hand appearance with synthetic annotations. Relying on image-to-image translation, we improve the appearance of synthetic hands to approximate the statistical distribution underlying a collection of real images of hands. H-GAN tackles not only the cross-domain tone mapping but also structural differences in localized areas such as shading discontinuities. Results are evaluated on a qualitative and quantitative basis improving previous works. Furthermore, we relied on the hand classification task to claim our generated hands are statistically similar to the real domain of hands.

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。

↑