arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2003.08791cs.CVcs.LGeess.IV
收藏

高分辨率日间转换(无需域标签)

High-Resolution Daytime Translation Without Domain Labels

Samsung AI Center, Moscow(三星人工智能中心(莫斯科));National Research University Higher School of Economics, St.-Petersburg(国立研究型大学高等经济学院(圣彼得堡));Skolkovo Institute of Science and Technology, Moscow(斯科尔科沃科学技术研究院(莫斯科))

Ivan Anokhin, Pavel Solovev, Denis Korzhenkov, Alexey Kharlamov, Taras Khakhulin, Alexey Silvestrov, Sergey Nikolenko, Victor Lempitsky, Gleb Sterkin

AI总结:

提出HiDT模型,结合生成式图像转换与新型上采样方案,实现无标签高分辨率日间转换,性能优越。

摘要:

对高分辨率照片中的日间变化进行建模,例如在白天、夜晚或黎明等不同光照条件下重新渲染同一场景,是一项具有挑战性的图像处理任务。我们针对该任务提出了高分辨率日间转换(HiDT)模型。HiDT结合了生成式图像到图像模型与一种新的上采样方案,使得图像转换能够在高分辨率下应用。该模型在常用的GAN指标和人工评估方面均表现出有竞争力的结果。重要的是,这种良好的性能是在仅使用无日间标签的静态风景图像数据集进行训练的情况下取得的。我们的结果可在https://saic-mdal.github.io/HiDT/获取。

英文摘要:

Modeling daytime changes in high resolution photographs, e.g., re-rendering the same scene under different illuminations typical for day, night, or dawn, is a challenging image manipulation task. We present the high-resolution daytime translation (HiDT) model for this task. HiDT combines a generative image-to-image model and a new upsampling scheme that allows to apply image translation at high resolution. The model demonstrates competitive results in terms of both commonly used GAN metrics and human evaluation. Importantly, this good performance comes as a result of training on a dataset of still landscape images with no daytime labels available. Our results are available at https://saic-mdal.github.io/HiDT/.

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。

↑