BriefGPT.xyz
May, 2024
基于特征融合网络的人机可扩展图像编码
Scalable Image Coding for Humans and Machines Using Feature Fusion Network
HTML
PDF
Takahiro Shindo, Taiju Watanabe, Yui Tatsumi, Hiroshi Watanabe
TL;DR
我们提出了一种基于学习的可扩展图像编码方法,适用于多种图像识别模型。通过将机器的图像压缩模型与人类的压缩模型相结合,利用特征融合网络实现了高效的图像压缩,并且减少了参数的数量。通过评估图像压缩性能,我们证明了这种可扩展编码方法的有效性。
Abstract
As
image recognition models
become more prevalent, scalable coding methods for machines and humans gain more importance. Applications of
image recognition models
include traffic monitoring and farm management. In
→