ABSTRACT
Joint imag
联合 现有方
Image f
带有引已成功应用于
Severa
已经提
To addres
为了解决
In this paper,
在本文中,
We make the following contributions in this paper:
– We propose a learning-based framework for constructing joint image filters. Our network takes both target and guidance images into consideration and naturally handles the inconsistent structure problem.
– With the learned joint filter, we demonstrate the state-of-the-art performance on four joint depth upsampling datasets.
– We show that the model trained on the RGB/D dataset generalizes well to handle image data in a variety of domains.
我们在本文中做出了以下贡献:
– 我们提出了一个基于学习的框架来构建联合图像过滤器。 我们的网络同时考虑了目标图像和引导图像,并自然地处理了结构不一致的问题。
– 通过学习的联合滤波器,我们展示了四个联合深度上采样数据集的最新性能。
– 我们表明,在 RGB/D 数据集上训练的模型可以很好地泛化以处理各种领域的图像数据。
2、Related Work
Joint Image Filters
Joint
联合图
Seco
其次
Deep Models for Low-Level Vision.
While
虽然
3、Learning Deep Joint Image Filters
Our
我们
3.1 Network Architecture Design
To design
设计联合滤波器
We attribute
局限性归因于
We adop
我们采用
3.2 Relationship to Prior Work
The proposed framework
所提出的框架