Unsupervised Image-to-Image Translation Networks --- Reading Writing
Unsupervised Image-to-Image Translation Networks --- Reading Writing
2017.03.03
Motivations: most existing image to image translation algorithms are all need image pairs as training data for deep neural network, such as CGANs or VAEs. But in some cases, it is rather difficult to collect such training data. For example, the night and day image pairs, the perfect aligned thermal RGB image pairs, or sunning rainning, fogging, et al, which provide us a new challenging problem:
How to do image to image translation in a unsupervised fashion which do not need aligned image pairs ?
This paper proposed the UNIT framework (UNsupervised Image-to-image Translation network) to deal with this problem which combine VAE and GANs. The whole framework can be described as the following figures which seems complex but rather easy to understand.
There are two most important assumptions about the proposed framework:
1. we assume that the relationship between X1 and X2 does not only exist at the image level but also at the level of local patches or regions.
2. for any given images x1 and x2, there exists a common underlying representation z, such that we can cover both images from this underlying representation from each of the two input images.
VAEs: the encoder-generator pair {E1, G1} constitutes a VAE for the X1 domain, termed VAE1. Another pair of {E2, G2} constitutes a VAE for the X2 domain VAE2.
Weight-sharing : we enforce a weight-sharing constraint to relate the representations in the two VAEs.
GANs : two GANs are used to output the two domains.
Experiments:
Unsupervised Image-to-Image Translation Networks --- Reading Writing的更多相关文章
- Learning to Compare Image Patches via Convolutional Neural Networks --- Reading Summary
Learning to Compare Image Patches via Convolutional Neural Networks --- Reading Summary 2017.03.08 ...
- Unsupervised Image-to-Image Translation Networks
Abstract: 无监督图像到图像的翻译目的是学习不同域图像的一个联合分布,通过使用来自单独域图像的边缘分布.给定一个边缘分布,可以得到很多种联合分布.如果不加入额外的假设条件的话,从边缘分布无法推 ...
- reading/writing files in Python
file types: plaintext files, such as .txt .py Binary files, such as .docx, .pdf, iamges, spreadsheet ...
- 【Deep Learning】Hinton. Reducing the Dimensionality of Data with Neural Networks Reading Note
2006年,机器学习泰斗.多伦多大学计算机系教授Geoffery Hinton在Science发表文章,提出基于深度信念网络(Deep Belief Networks, DBN)可使用非监督的逐层贪心 ...
- On Explainability of Deep Neural Networks
On Explainability of Deep Neural Networks « Learning F# Functional Data Structures and Algorithms is ...
- The Unreasonable Effectiveness of Recurrent Neural Networks (RNN)
http://karpathy.github.io/2015/05/21/rnn-effectiveness/ There’s something magical about Recurrent Ne ...
- 26 THINGS I LEARNED IN THE DEEP LEARNING SUMMER SCHOOL
26 THINGS I LEARNED IN THE DEEP LEARNING SUMMER SCHOOL In the beginning of August I got the chance t ...
- DotNet 资源大全中文版(Awesome最新版)
Awesome系列的.Net资源整理.awesome-dotnet是由quozd发起和维护.内容包括:编译器.压缩.应用框架.应用模板.加密.数据库.反编译.IDE.日志.风格指南等. 算法与数据结构 ...
- (转) Awesome - Most Cited Deep Learning Papers
转自:https://github.com/terryum/awesome-deep-learning-papers Awesome - Most Cited Deep Learning Papers ...
随机推荐
- Java基础语法(三)
七.方法 定义: 方法就是完成特定功能的代码块 在很多语言里面都有函数的定义 函数在Java中被称为方法 格式: 修饰符 返回值类型 方法名(参数类型 参数名1,参数类型 参数名2…) { 函数体; ...
- CocoaPod 使用(I)
1. 添加一个 Podfile 文件 在终端先找到你要放入 Podfile 的文件路径: cd 文件路径 cd /Users/XXX/Desktop/RAC 然后敲入: pod init 就会给你建立 ...
- Qt 事件机制
[1]事件 事件是可以被控件识别的操作.如按下确定按钮.选择某个单选按钮或复选框. 每种控件有自己可识别的事件,如窗体的加载.单击.双击等事件,编辑框(文本框)的文本改变事件等等. 事件就是用户对窗口 ...
- JSP FreeMarker Velocity 原理
JSP原理 JSP的运行原理:JSP 本质上是一个Servlet. 每个JSP 页面在第一次被访问时,JSP引擎将它翻译成一个Servlet 程序,然后再把这个 Servlet 源程序编译成Servl ...
- JAVA 中的 Collection 和 Map 以及相关派生类的概念
JAVA中Collection接口和Map接口的主要实现类 Collection接口 Collection是最基本的集合接口,一个Collection代表一组Object,即Collection的 ...
- webpack4.0 实战记录
从零配置webpack4.0 搭建React工程. 基本环境:Node(v8.1.2)+ webpack(v4.16.2) 1.在项目目录 命令窗口 执行 npm init 初始化项目,执行完后项 ...
- C#获取驱动器盘符
一.使用ManagementObjectSearcher类 static void Main(string[] args) { SelectQuery selectQuery = new Select ...
- HDU 1207 汉诺塔II (递推)
经典的汉诺塔问题经常作为一个递归的经典例题存在.可能有人并不知道汉诺塔问题的典故.汉诺塔来源于印度传说的一个故事,上帝创造世界时作了三根金刚石柱子,在一根柱子上从下往上按大小顺序摞着64片黄金圆盘.上 ...
- [转载]转,Oracle中关于处理小数点位数的几个函数,取小数位数,Oracle查询函数
关于处理小数点位数的几个oracle函数() 1. 取四舍五入的几位小数 select round(1.2345, 3) from dual; 结果:1.235 2. 保留两位小数,只舍 select ...
- FFMPEG结构体分析:AVFrame(解码后的数据)
https://blog.csdn.net/jxcr1984/article/details/52766524 本文转自: http://blog.csdn.net/leixiaohua1020/ar ...