arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2008.03706cs.CVeess.IV
收藏

块洗牌:一种在有限内存下实现高分辨率快速风格转换的方法

Block Shuffle: A Method for High-resolution Fast Style Transfer with Limited Memory

Weifeng Ma, Zhe Chen, Caoting Ji

摘要:

快速风格转换是一系列使用前馈神经网络渲染输入图像的神经风格转换算法。由于输出层的维数很高,这些网络需要大量的内存来进行计算。因此,对于高分辨率的图像,大多数移动设备和个人电脑都无法对其进行风格转换,这极大地限制了快速风格传输的应用场景。目前,现有的两种解决方案是购买更多的内存和使用基于羽化的方法,但前者需要额外的成本,后者的图像质量较差。为了解决这个问题,我们提出了一种新的图像合成方法命名为\emph{block shuffle},它将一个高内存消耗的单个任务转换为多个子任务与低内存消耗。此方法可以作为快速样式传输的插件,而无需对网络架构进行任何修改。我们使用GitHub上最流行的快速样式传输存储库作为基线。实验结果表明,该方法生成的高分辨率图像质量优于基于羽化的方法。虽然我们的方法比基线慢一个数量级,但它可以在有限内存的情况下处理高分辨率图像,这在基线下是不可能的。代码和型号将在\url{https://github.com/czczup/block-shuffle}上提供。

英文摘要:

Fast Style Transfer is a series of Neural Style Transfer algorithms that use feed-forward neural networks to render input images. Because of the high dimension of the output layer, these networks require much memory for computation. Therefore, for high-resolution images, most mobile devices and personal computers cannot stylize them, which greatly limits the application scenarios of Fast Style Transfer. At present, the two existing solutions are purchasing more memory and using the feathering-based method, but the former requires additional cost, and the latter has poor image quality. To solve this problem, we propose a novel image synthesis method named \emph{block shuffle}, which converts a single task with high memory consumption to multiple subtasks with low memory consumption. This method can act as a plug-in for Fast Style Transfer without any modification to the network architecture. We use the most popular Fast Style Transfer repository on GitHub as the baseline. Experiments show that the quality of high-resolution images generated by our method is better than that of the feathering-based method. Although our method is an order of magnitude slower than the baseline, it can stylize high-resolution images with limited memory, which is impossible with the baseline. The code and models will be made available on \url{https://github.com/czczup/block-shuffle}.

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。

↑