BriefGPT.xyz
Dec, 2018
JPEG转换域中的深度残差学习
Deep Residual Learning in the JPEG Transform Domain
HTML
PDF
Max Ehrlich, Larry Davis
TL;DR
利用JPEG变换域的线性特性重新定义了卷积和批量归一化，实现在压缩图像作为输入的Residual Network推断和学习，无损准确性地进行图像分类。
Abstract
We introduce a general method of performing
residual network
inference and learning in the
jpeg transform domain
that allows the network to consume compressed images as input. Our formulation leverages the linear
→