可逆神经网络的研究及其在图像中应用

一、摘要

可逆神经网络(INN)自被提出以来，就受到了广泛关注。由于其双射构造和高效可逆性，INN被用于各种推理任务，如图像隐藏、图像重缩放、图像着色、图像压缩和视频超分辨率等等。本文针对最新关于INN在图像方面应用的文献进行介绍，包括每篇文献的基本原理和个人理解。最后对所介绍的文献进行总结，指出各自的优缺点并对未来INN在图像方面应用展开思考。

Invertible neural network (INN) has been widely concerned since it was proposed. Because of its bijective construction and efficient reversibility, INN is used for a variety of inference tasks, such as image hiding, image rescaling, image coloring, image compression, and video super resolution, among others. This article introduces the latest literature on the use of INN in imaging, including the rationale and personal understanding of each literature. Finally, the paper summarizes the literature, points out their advantages a