BriefGPT.xyz
Feb, 2019
深度高光谱先验: 降噪、修复、超分辨率
Deep Hyperspectral Prior: Denoising, Inpainting, Super-Resolution
HTML
PDF
Oleksii Sidorov, Jon Yngve Hardeberg
TL;DR
本篇研究旨在通过利用CNN的内在属性来解决高光谱图像处理中数据集过小导致训练困难的问题，并提出了一种新的不需训练网络的方法解决高光谱图像去噪、修复和超分辨率问题，该方法在性能上与已有的训练网络方法相当，但不受训练数据可用性限制。
Abstract
deep learning
algorithms have demonstrated state-of-the-art performance in various tasks of image restoration. This was made possible through the ability of
cnns
to learn from large exemplar sets. However, the la
→