<?xml version="1.0" encoding="utf-8" ?><rss version="2.0"><channel><title><![CDATA[xihang_alpha的博客]]></title><description><![CDATA[]]></description><link>https://blog.youkuaiyun.com/xihang_alpha</link><language>zh-cn</language><generator>https://blog.youkuaiyun.com/</generator><copyright><![CDATA[Copyright &copy; xihang_alpha]]></copyright><item><title><![CDATA[Network in Network]]></title><link>https://blog.youkuaiyun.com/xihang_alpha/article/details/70195031</link><guid>https://blog.youkuaiyun.com/xihang_alpha/article/details/70195031</guid><author>xihang_alpha</author><pubDate>Sun, 16 Apr 2017 13:19:04 +0800</pubDate><description><![CDATA[一.文献名字和作者
     Network In Network.Min Lin, Qiang Chen, Shuicheng Yan
   
二.阅读时间
    2014年11月19日
三.文献的目的
    文章主要是对于CNN结构进行改进，使得CNN能够学习到更加抽象和有效的非线性特征。
四.文献的贡献点
4.1 卷积层的改进–多层感知卷积层
    传统的卷积层只是将前一层的特征进行了线]]></description><category></category></item><item><title><![CDATA[（Paper）Network in Network网络分析]]></title><link>https://blog.youkuaiyun.com/xihang_alpha/article/details/70194936</link><guid>https://blog.youkuaiyun.com/xihang_alpha/article/details/70194936</guid><author>xihang_alpha</author><pubDate>Sun, 16 Apr 2017 12:59:35 +0800</pubDate><description><![CDATA[本文地址：http://blog.youkuaiyun.com/mounty_fsc/article/details/51746111
《Network in Network》论文笔记1.综述这篇文章有两个很重要的观点：
1×1卷积的使用 
文中提出使用mlpconv网络层替代传统的convolution层。mlp层实际上是卷积加传统的mlp（多层感知器），因为convolution是线性的，而mlp是非线性的]]></description><category></category></item><item><title><![CDATA[2014-Network In Network]]></title><link>https://blog.youkuaiyun.com/xihang_alpha/article/details/70194724</link><guid>https://blog.youkuaiyun.com/xihang_alpha/article/details/70194724</guid><author>xihang_alpha</author><pubDate>Sun, 16 Apr 2017 12:20:54 +0800</pubDate><description><![CDATA[Network In Network学习笔记原文地址：http://blog.youkuaiyun.com/hjimce/article/details/50458190作者：hjimce一、相关理论本篇博文主要讲解2014年ICLR的一篇非常牛逼的paper：《Network In Network》，过去一年已经有了好几百的引用量，这篇paper改进了传统的CNN网络，采用了少量的参数就松松击败了Alexne]]></description><category></category></item><item><title><![CDATA[ALEXNET解读]]></title><link>https://blog.youkuaiyun.com/xihang_alpha/article/details/70194204</link><guid>https://blog.youkuaiyun.com/xihang_alpha/article/details/70194204</guid><author>xihang_alpha</author><pubDate>Sun, 16 Apr 2017 10:56:33 +0800</pubDate><description><![CDATA[转自csdn blog这篇博客讲现在很流行的两种网络模型，ResNet和DenseNet,其实可以把DenseNet看做是ResNet的特例 
文章地址: 
[1]Deep Residual Learning for Image Recognition,CVPR2015 
[2]Densely Connected Convolutional Networks,CVPR2016本篇博客不讲论文的内容，]]></description><category></category></item><item><title><![CDATA[ResNet,DenseNet,以及残差家族]]></title><link>https://blog.youkuaiyun.com/xihang_alpha/article/details/70194074</link><guid>https://blog.youkuaiyun.com/xihang_alpha/article/details/70194074</guid><author>xihang_alpha</author><pubDate>Sun, 16 Apr 2017 10:35:39 +0800</pubDate><description><![CDATA[转自：http://blog.youkuaiyun.com/cv_family_z/article/details/50328175CVPR2016 
https://github.com/KaimingHe/deep-residual-networks
这是微软方面的最新研究成果， 在第六届ImageNet年度图像识别测试中，微软研究院的计算机图像识别系统在几个类别的测试中获得第一名。
本文是解决超深度CNN]]></description><category></category></item><item><title><![CDATA[Re-thinking Deep Residual Networks]]></title><link>https://blog.youkuaiyun.com/xihang_alpha/article/details/70187590</link><guid>https://blog.youkuaiyun.com/xihang_alpha/article/details/70187590</guid><author>xihang_alpha</author><pubDate>Sat, 15 Apr 2017 21:55:20 +0800</pubDate><description><![CDATA[本文是对ImageNet 2015的冠军ResNet(Deep Residual Networks)以及目前围绕ResNet这个工作研究者后续所发论文的总结，主要涉及到下面5篇论文。]]></description><category></category></item><item><title><![CDATA[CNN经典模型总结]]></title><link>https://blog.youkuaiyun.com/xihang_alpha/article/details/70186973</link><guid>https://blog.youkuaiyun.com/xihang_alpha/article/details/70186973</guid><author>xihang_alpha</author><pubDate>Sat, 15 Apr 2017 21:03:41 +0800</pubDate><description><![CDATA[转自：http://www.open-open.com/lib/view/open1473213789568.html
参考：http://blog.youkuaiyun.com/qq_17754181/article/details/59111864?locationNum=14&fps=1
略有删改。
LeNet5LeNet5 诞生于 1994 年，是最早的卷积神经网络之一，并且推动了深度学习领域的发展。自]]></description><category></category></item><item><title><![CDATA[Convergent Learning: Do different neural networks learn the same representations?]]></title><link>https://blog.youkuaiyun.com/xihang_alpha/article/details/70186760</link><guid>https://blog.youkuaiyun.com/xihang_alpha/article/details/70186760</guid><author>xihang_alpha</author><pubDate>Sat, 15 Apr 2017 20:39:07 +0800</pubDate><description><![CDATA[Jason Yosinski's lecture on <<Convergent Learning: Do different neural networks learn the same representations?>>]]></description><category></category></item><item><title><![CDATA[2016-Deep Learning Trends @ICLR]]></title><link>https://blog.youkuaiyun.com/xihang_alpha/article/details/70186430</link><guid>https://blog.youkuaiyun.com/xihang_alpha/article/details/70186430</guid><author>xihang_alpha</author><pubDate>Sat, 15 Apr 2017 19:43:04 +0800</pubDate><description><![CDATA[Cited fromTombone’s Computer Vision Blog
Deep Learning, Computer Vision, and the algorithms that are shaping the future of Artificial Intelligence.
 
 
  
 
          Wednesday, June 01, 2016]]></description><category></category></item><item><title><![CDATA[转载博客文章技巧]]></title><link>https://blog.youkuaiyun.com/xihang_alpha/article/details/70186356</link><guid>https://blog.youkuaiyun.com/xihang_alpha/article/details/70186356</guid><author>xihang_alpha</author><pubDate>Sat, 15 Apr 2017 19:27:18 +0800</pubDate><description><![CDATA[对于喜欢逛优快云的人来说，看别人的博客确实能够对自己有不小的提高，有时候看到特别好的博客想转载下载，但是不能一个字一个字的敲了，这时候我们就想快速转载别人的博客，把别人的博客移到自己的空间里面，当然有人会说我们可以收藏博客啊，就不需要转载，(⊙o⊙)… 也对。。实现  因为我自己当初想转载的时候却不知道该怎么转载，所以学会了之后就把方法写出来，帮助那些想转载却不知道该怎么转载的人（大神勿笑）。]]></description><category></category></item></channel></rss>