人工智能培训

搜索

深度学习论文:基于深度特征重建的可扩展面部图像压缩(Scalable Facial Image Compression with Deep Feature Re

[复制链接]
redfrog 发表于 2019-3-15 12:05:55 | 显示全部楼层 |阅读模式
redfrog 2019-3-15 12:05:55 698 0 显示全部楼层
深度学习论文:基于深度特征重建的可扩展面部图像压缩(Scalable Facial Image Compression with Deep Feature Reconstruction)在本文中,我们提出了一种可伸缩的图像压缩方案,包括用于特征表示的基础层和用于纹理表示的增强层。更具体地说,基础层被设计为用于分析目的的加深学习特征,并且还可以通过深度特征重建将其转换为细化结构。增强层用于压缩输入图像和从基层产生的信号之间的残差,旨在忠实地重建输入纹理。所提出的方案可以在监控应用中可行地继承压缩 - 分析和分析 - 然后 - 压缩方案的优点。该框架的性能通过面部图像验证,并且所进行的实验提供了有用的证据,以表明所提出的框架可以实现比传统图像压缩方案更好的速率准确性和速率 - 失真性能。
In this paper, we propose a scalable image compression scheme, including thebase layer for feature representation and enhancement layer for texturerepresentation.More specifically, the base layer is designed as the deeplearning feature for analysis purpose, and it can also be converted to the finestructure with deep feature reconstruction.The enhancement layer, which servesto compress the residuals between the input image and the signals generatedfrom the base layer, aims to faithfully reconstruct the input texture.Theproposed scheme can feasibly inherit the advantages of bothcompress-then-analyze and analyze-then-compress schemes in surveillanceapplications.The performance of this framework is validated with facialimages, and the conducted experiments provide useful evidences to show that theproposed framework can achieve better rate-accuracy and rate-distortionperformance over conventional image compression schemes.深度学习论文:基于深度特征重建的可扩展面部图像压缩(Scalable Facial Image Compression with Deep Feature Reconstruction) jqQd6xY3IFfTgirM.jpg
URL地址:https://arxiv.org/abs/1903.05921     ----pdf下载地址:https://arxiv.org/pdf/1903.05921    ----深度学习论文:基于深度特征重建的可扩展面部图像压缩(Scalable Facial Image Compression with Deep Feature Reconstruction)
回复

使用道具 举报

您需要登录后才可以回帖 登录 | 立即注册

本版积分规则 返回列表 发新帖

redfrog当前离线
新手上路

查看:698 | 回复:0

快速回复 返回顶部 返回列表