atom.xml

<?xml version="1.0" encoding="utf-8"?>
<feed xmlns="http://www.w3.org/2005/Atom">
  <title>lvskiller&#39;s Blog</title>
  
  
  <link href="/atom.xml" rel="self"/>
  
  <link href="http://yoursite.com/"/>
  <updated>2022-01-09T15:46:43.757Z</updated>
  <id>http://yoursite.com/</id>
  
  <author>
    <name>lvskiller</name>
    
  </author>
  
  <generator uri="https://hexo.io/">Hexo</generator>
  
  <entry>
    <title>linux安装openCV4.5.x</title>
    <link href="http://yoursite.com/2022/01/09/linux%E5%AE%89%E8%A3%85openCV4-5-x/"/>
    <id>http://yoursite.com/2022/01/09/linux%E5%AE%89%E8%A3%85openCV4-5-x/</id>
    <published>2022-01-09T09:15:07.000Z</published>
    <updated>2022-01-09T15:46:43.757Z</updated>
    
    <summary type="html">
    
      &lt;h1 id=&quot;linux安装openCV4-5-x&quot;&gt;&lt;a href=&quot;#linux安装openCV4-5-x&quot; class=&quot;headerlink&quot; title=&quot;linux安装openCV4.5.x&quot;&gt;&lt;/a&gt;linux安装openCV4.5.x&lt;/h1&gt;&lt;blockquote&gt;
&lt;p&gt;这次总算是sudo了，所以直接一套操作猛如虎了。&lt;/p&gt;
&lt;/blockquote&gt;
    
    </summary>
    
    
    
  </entry>
  
  <entry>
    <title>非root降低gcc版本到5.4</title>
    <link href="http://yoursite.com/2022/01/09/%E9%99%8D%E4%BD%8Egcc%E7%89%88%E6%9C%AC%E5%88%B05-4/"/>
    <id>http://yoursite.com/2022/01/09/%E9%99%8D%E4%BD%8Egcc%E7%89%88%E6%9C%AC%E5%88%B05-4/</id>
    <published>2022-01-09T08:26:47.000Z</published>
    <updated>2022-01-09T09:11:40.104Z</updated>
    
    <summary type="html">
    
      &lt;h1 id=&quot;降低gcc版本到5-4&quot;&gt;&lt;a href=&quot;#降低gcc版本到5-4&quot; class=&quot;headerlink&quot; title=&quot;降低gcc版本到5.4&quot;&gt;&lt;/a&gt;降低gcc版本到5.4&lt;/h1&gt;&lt;h2 id=&quot;非root降低gcc版本到5-4&quot;&gt;&lt;a href=&quot;#非root降低gcc版本到5-4&quot; class=&quot;headerlink&quot; title=&quot;非root降低gcc版本到5.4&quot;&gt;&lt;/a&gt;非root降低gcc版本到5.4&lt;/h2&gt;&lt;blockquote&gt;
&lt;p&gt;还是那个MAttNet惹的祸，它里面的东西的make需要gcc 5.4.0，人给他整麻了&lt;/p&gt;
&lt;/blockquote&gt;
    
    </summary>
    
    
      <category term="linux" scheme="http://yoursite.com/categories/linux/"/>
    
    
  </entry>
  
  <entry>
    <title>非sudo权限将cuda降版本</title>
    <link href="http://yoursite.com/2022/01/09/cuda10-to-cuda8/"/>
    <id>http://yoursite.com/2022/01/09/cuda10-to-cuda8/</id>
    <published>2022-01-09T06:57:51.000Z</published>
    <updated>2022-01-09T08:20:12.380Z</updated>
    
    <summary type="html">
    
      &lt;h2 id=&quot;非sudo权限将cuda降版本&quot;&gt;&lt;a href=&quot;#非sudo权限将cuda降版本&quot; class=&quot;headerlink&quot; title=&quot;非sudo权限将cuda降版本&quot;&gt;&lt;/a&gt;非sudo权限将cuda降版本&lt;/h2&gt;&lt;h1 id=&quot;全连接神经网络&quot;&gt;&lt;a href=&quot;#全连接神经网络&quot; class=&quot;headerlink&quot; title=&quot;全连接神经网络&quot;&gt;&lt;/a&gt;全连接神经网络&lt;/h1&gt;&lt;blockquote&gt;
&lt;p&gt;背景： 主要就是MAttNet它需要的torch版本是0.2+，这个版本的cuda是8.0，显然现在大部分都是11了，所以需要非sudo权限降低版本&lt;/p&gt;
&lt;/blockquote&gt;
    
    </summary>
    
    
      <category term="linux" scheme="http://yoursite.com/categories/linux/"/>
    
    
  </entry>
  
  <entry>
    <title>complie theory</title>
    <link href="http://yoursite.com/2021/05/26/complie-theory/"/>
    <id>http://yoursite.com/2021/05/26/complie-theory/</id>
    <published>2021-05-26T07:45:06.000Z</published>
    <updated>2021-05-26T07:45:06.419Z</updated>
    
    <summary type="html">
    
      &lt;h1 id=&quot;基本语句&quot;&gt;&lt;a href=&quot;#基本语句&quot; class=&quot;headerlink&quot; title=&quot;基本语句&quot;&gt;&lt;/a&gt;基本语句&lt;/h1&gt;&lt;blockquote&gt;
&lt;p&gt;学习论文：Semantic Image Synthesis with Spatially-Adaptive Normalization&lt;/p&gt;
&lt;p&gt;论文目的：It propose a new normalization layer to solve the problem of original normalization.&lt;/p&gt;
&lt;p&gt;关键词： &lt;strong&gt;Spatially-Adaptive Normalization&lt;/strong&gt;&lt;/p&gt;
&lt;/blockquote&gt;
    
    </summary>
    
    
    
  </entry>
  
  <entry>
    <title>pytorch_tutial</title>
    <link href="http://yoursite.com/2021/05/17/pytorch-tutial/"/>
    <id>http://yoursite.com/2021/05/17/pytorch-tutial/</id>
    <published>2021-05-17T07:49:27.000Z</published>
    <updated>2021-05-17T07:49:27.985Z</updated>
    
    <summary type="html">
    
      &lt;h1 id=&quot;基本语句&quot;&gt;&lt;a href=&quot;#基本语句&quot; class=&quot;headerlink&quot; title=&quot;基本语句&quot;&gt;&lt;/a&gt;基本语句&lt;/h1&gt;&lt;blockquote&gt;
&lt;p&gt;学习论文：Semantic Image Synthesis with Spatially-Adaptive Normalization&lt;/p&gt;
&lt;p&gt;论文目的：It propose a new normalization layer to solve the problem of original normalization.&lt;/p&gt;
&lt;p&gt;关键词： &lt;strong&gt;Spatially-Adaptive Normalization&lt;/strong&gt;&lt;/p&gt;
&lt;/blockquote&gt;
    
    </summary>
    
    
    
  </entry>
  
  <entry>
    <title>Shell-1</title>
    <link href="http://yoursite.com/2021/04/29/Shell-1/"/>
    <id>http://yoursite.com/2021/04/29/Shell-1/</id>
    <published>2021-04-29T00:00:04.000Z</published>
    <updated>2021-05-02T12:18:26.224Z</updated>
    
    <summary type="html">
    
      &lt;h1 id=&quot;Shell-1&quot;&gt;&lt;a href=&quot;#Shell-1&quot; class=&quot;headerlink&quot; title=&quot;Shell-1&quot;&gt;&lt;/a&gt;Shell-1&lt;/h1&gt;&lt;blockquote&gt;
&lt;p&gt;学习linux的shell操作&lt;/p&gt;
&lt;/blockquote&gt;
    
    </summary>
    
    
    
  </entry>
  
  <entry>
    <title>CSP-前缀和</title>
    <link href="http://yoursite.com/2021/02/12/CSP-%E5%89%8D%E7%BC%80%E5%92%8C/"/>
    <id>http://yoursite.com/2021/02/12/CSP-%E5%89%8D%E7%BC%80%E5%92%8C/</id>
    <published>2021-02-12T08:33:38.000Z</published>
    <updated>2021-02-12T08:33:38.748Z</updated>
    
    <summary type="html">
    
      &lt;h1 id=&quot;基本语句&quot;&gt;&lt;a href=&quot;#基本语句&quot; class=&quot;headerlink&quot; title=&quot;基本语句&quot;&gt;&lt;/a&gt;基本语句&lt;/h1&gt;&lt;blockquote&gt;
&lt;p&gt;学习论文：Semantic Image Synthesis with Spatially-Adaptive Normalization&lt;/p&gt;
&lt;p&gt;论文目的：It propose a new normalization layer to solve the problem of original normalization.&lt;/p&gt;
&lt;p&gt;关键词： &lt;strong&gt;Spatially-Adaptive Normalization&lt;/strong&gt;&lt;/p&gt;
&lt;/blockquote&gt;
    
    </summary>
    
    
    
  </entry>
  
  <entry>
    <title>ALAE</title>
    <link href="http://yoursite.com/2021/02/09/ALAE/"/>
    <id>http://yoursite.com/2021/02/09/ALAE/</id>
    <published>2021-02-09T11:27:07.000Z</published>
    <updated>2021-02-09T11:27:07.338Z</updated>
    
    <summary type="html">
    
      &lt;h1 id=&quot;基本语句&quot;&gt;&lt;a href=&quot;#基本语句&quot; class=&quot;headerlink&quot; title=&quot;基本语句&quot;&gt;&lt;/a&gt;基本语句&lt;/h1&gt;&lt;blockquote&gt;
&lt;p&gt;学习论文：Semantic Image Synthesis with Spatially-Adaptive Normalization&lt;/p&gt;
&lt;p&gt;论文目的：It propose a new normalization layer to solve the problem of original normalization.&lt;/p&gt;
&lt;p&gt;关键词： &lt;strong&gt;Spatially-Adaptive Normalization&lt;/strong&gt;&lt;/p&gt;
&lt;/blockquote&gt;
    
    </summary>
    
    
    
  </entry>
  
  <entry>
    <title>StarGAN V2</title>
    <link href="http://yoursite.com/2021/02/08/StarGAN-V2/"/>
    <id>http://yoursite.com/2021/02/08/StarGAN-V2/</id>
    <published>2021-02-08T04:57:51.000Z</published>
    <updated>2021-02-08T12:15:27.934Z</updated>
    
    <summary type="html">
    
      &lt;h1 id=&quot;StarGAN-V2&quot;&gt;&lt;a href=&quot;#StarGAN-V2&quot; class=&quot;headerlink&quot; title=&quot;StarGAN V2&quot;&gt;&lt;/a&gt;StarGAN V2&lt;/h1&gt;&lt;blockquote&gt;
&lt;p&gt;学习论文：StarGAN v2: Diverse Image Synthesis for Multiple Domains&lt;/p&gt;
&lt;p&gt;论文目的：It propose a new version of StarGAN that can generate diverse images and scalability over mutiple domains.&lt;/p&gt;
&lt;p&gt;关键词： &lt;strong&gt;Diversity&lt;/strong&gt;    &lt;strong&gt;Multi-Domain&lt;/strong&gt;&lt;/p&gt;
&lt;/blockquote&gt;
    
    </summary>
    
    
    
  </entry>
  
  <entry>
    <title>CSP-递推和递归</title>
    <link href="http://yoursite.com/2021/02/08/CSP-%E9%80%92%E6%8E%A8%E5%92%8C%E9%80%92%E5%BD%92/"/>
    <id>http://yoursite.com/2021/02/08/CSP-%E9%80%92%E6%8E%A8%E5%92%8C%E9%80%92%E5%BD%92/</id>
    <published>2021-02-08T03:01:00.000Z</published>
    <updated>2021-02-12T08:32:36.923Z</updated>
    
    <summary type="html">
    
      &lt;h1 id=&quot;CSP-递推和递归&quot;&gt;&lt;a href=&quot;#CSP-递推和递归&quot; class=&quot;headerlink&quot; title=&quot;CSP-递推和递归&quot;&gt;&lt;/a&gt;CSP-递推和递归&lt;/h1&gt;&lt;blockquote&gt;
&lt;p&gt;该篇介绍了常见的递推和递归想法&lt;/p&gt;
&lt;/blockquote&gt;
    
    </summary>
    
    
      <category term="CSP" scheme="http://yoursite.com/categories/CSP/"/>
    
    
  </entry>
  
  <entry>
    <title>Javax初学</title>
    <link href="http://yoursite.com/2021/01/30/Javax%E5%88%9D%E5%AD%A6/"/>
    <id>http://yoursite.com/2021/01/30/Javax%E5%88%9D%E5%AD%A6/</id>
    <published>2021-01-30T10:31:17.000Z</published>
    <updated>2021-01-30T12:52:21.965Z</updated>
    
    <summary type="html">
    
      &lt;h1 id=&quot;Java初学&quot;&gt;&lt;a href=&quot;#Java初学&quot; class=&quot;headerlink&quot; title=&quot;Java初学&quot;&gt;&lt;/a&gt;Java初学&lt;/h1&gt;&lt;blockquote&gt;
&lt;p&gt;该篇介绍了Java的基本语句&lt;/p&gt;
&lt;/blockquote&gt;
    
    </summary>
    
    
    
  </entry>
  
  <entry>
    <title>FaseShifer论文精读</title>
    <link href="http://yoursite.com/2021/01/26/FaseShifer%E8%AE%BA%E6%96%87%E7%B2%BE%E8%AF%BB/"/>
    <id>http://yoursite.com/2021/01/26/FaseShifer%E8%AE%BA%E6%96%87%E7%B2%BE%E8%AF%BB/</id>
    <published>2021-01-26T09:35:45.000Z</published>
    <updated>2021-02-10T11:59:10.957Z</updated>
    
    <summary type="html">
    
      &lt;h1 id=&quot;FaseShifter论文精读&quot;&gt;&lt;a href=&quot;#FaseShifter论文精读&quot; class=&quot;headerlink&quot; title=&quot;FaseShifter论文精读&quot;&gt;&lt;/a&gt;FaseShifter论文精读&lt;/h1&gt;&lt;blockquote&gt;
&lt;p&gt;学习论文：FaceShifter: Towards High Fidelity And Occlusion Aware Face Swapping&lt;/p&gt;
&lt;p&gt;论文目的：This paper proposes a two-staged method callded FaceShifer to get high quality Face Swap.AEI-net make full use of the information from the targert image.And Hear-net can explozit the Fidelity and occlusion.&lt;/p&gt;
&lt;p&gt;关键词：&lt;strong&gt;Face Swap&lt;/strong&gt;   &lt;/p&gt;
&lt;/blockquote&gt;
    
    </summary>
    
    
    
  </entry>
  
  <entry>
    <title>Perceptual-Losses</title>
    <link href="http://yoursite.com/2021/01/24/Perceptual-Losses/"/>
    <id>http://yoursite.com/2021/01/24/Perceptual-Losses/</id>
    <published>2021-01-24T12:12:07.000Z</published>
    <updated>2021-01-26T09:25:14.341Z</updated>
    
    <summary type="html">
    
      &lt;h1 id=&quot;Perceptual-Losses&quot;&gt;&lt;a href=&quot;#Perceptual-Losses&quot; class=&quot;headerlink&quot; title=&quot;Perceptual Losses&quot;&gt;&lt;/a&gt;Perceptual Losses&lt;/h1&gt;&lt;blockquote&gt;
&lt;p&gt;学习论文：Perceptual Losses for Real-Time Style Transfer and Super-Resolution&lt;/p&gt;
&lt;p&gt;论文目的：This paper propose a loss(Maybe has parallel work，but this paper is much faster),called Perceptual Loss to take place with per-pixel loss,and it also get application in Style transfer, super-resolution。&lt;/p&gt;
&lt;p&gt;关键词：&lt;strong&gt;Style transfer&lt;/strong&gt;    &lt;strong&gt;super-resolution&lt;/strong&gt;    &lt;strong&gt;Perceptual Losses&lt;/strong&gt;&lt;/p&gt;
&lt;/blockquote&gt;
    
    </summary>
    
    
    
  </entry>
  
  <entry>
    <title>SPADE论文精读</title>
    <link href="http://yoursite.com/2021/01/23/SPADE%E8%AE%BA%E6%96%87%E7%B2%BE%E8%AF%BB/"/>
    <id>http://yoursite.com/2021/01/23/SPADE%E8%AE%BA%E6%96%87%E7%B2%BE%E8%AF%BB/</id>
    <published>2021-01-23T00:05:58.000Z</published>
    <updated>2021-01-23T08:48:28.839Z</updated>
    
    <summary type="html">
    
      &lt;h1 id=&quot;SPADE论文精读&quot;&gt;&lt;a href=&quot;#SPADE论文精读&quot; class=&quot;headerlink&quot; title=&quot;SPADE论文精读&quot;&gt;&lt;/a&gt;SPADE论文精读&lt;/h1&gt;&lt;blockquote&gt;
&lt;p&gt;学习论文：Semantic Image Synthesis with Spatially-Adaptive Normalization&lt;/p&gt;
&lt;p&gt;论文目的：It propose a new normalization layer to solve the problem of original normalization.&lt;/p&gt;
&lt;p&gt;论文摘要：To solve the normalization tend to “wash away” semantic information,the paper propose a new Spatially-Adaptive Normalization.And it also can control over both semantic and style.&lt;/p&gt;
&lt;p&gt;关键词： &lt;strong&gt;Spatially-Adaptive Normalization&lt;/strong&gt;&lt;/p&gt;
&lt;/blockquote&gt;
    
    </summary>
    
    
    
  </entry>
  
  <entry>
    <title>CSP-位运算</title>
    <link href="http://yoursite.com/2021/01/18/CSP-%E4%BD%8D%E8%BF%90%E7%AE%97/"/>
    <id>http://yoursite.com/2021/01/18/CSP-%E4%BD%8D%E8%BF%90%E7%AE%97/</id>
    <published>2021-01-18T13:25:41.000Z</published>
    <updated>2021-02-08T02:57:21.551Z</updated>
    
    <summary type="html">
    
      &lt;h1 id=&quot;CSP-位运算&quot;&gt;&lt;a href=&quot;#CSP-位运算&quot; class=&quot;headerlink&quot; title=&quot;CSP-位运算&quot;&gt;&lt;/a&gt;CSP-位运算&lt;/h1&gt;&lt;blockquote&gt;
&lt;p&gt;该篇介绍了介绍了位运算这种常用的算法以及使用位优化进行的算法。&lt;/p&gt;
&lt;/blockquote&gt;
    
    </summary>
    
    
    
  </entry>
  
  <entry>
    <title>pix2pix论文精读</title>
    <link href="http://yoursite.com/2021/01/18/pix2pix%E8%AE%BA%E6%96%87%E7%B2%BE%E8%AF%BB/"/>
    <id>http://yoursite.com/2021/01/18/pix2pix%E8%AE%BA%E6%96%87%E7%B2%BE%E8%AF%BB/</id>
    <published>2021-01-18T06:31:31.000Z</published>
    <updated>2021-01-22T14:35:22.639Z</updated>
    
    <summary type="html">
    
      &lt;h1 id=&quot;pix2pix论文精读&quot;&gt;&lt;a href=&quot;#pix2pix论文精读&quot; class=&quot;headerlink&quot; title=&quot;pix2pix论文精读&quot;&gt;&lt;/a&gt;pix2pix论文精读&lt;/h1&gt;&lt;blockquote&gt;
&lt;p&gt;学习论文：Image-to-Image Translation with Conditional Adversarial Networks&lt;/p&gt;
&lt;p&gt;论文目的：To achieve Image-to-image translation,it propose a new GAN callded Conditional Adversarial Networks(cGAN). &lt;/p&gt;
&lt;p&gt;论文摘要: Image-to-image translation is a important topic,such as  colorizing the picturces,reconstruct the objects from edge edges.And it also get many application.Our purpose is developing a common framwork to complete this problem instead of special-purpose machine.&lt;/p&gt;
&lt;p&gt;关键词：&lt;strong&gt;cGAN&lt;/strong&gt;     &lt;strong&gt;Image-to-image translation&lt;/strong&gt;&lt;/p&gt;
&lt;/blockquote&gt;
    
    </summary>
    
    
    
  </entry>
  
  <entry>
    <title>CycleGAN论文精读</title>
    <link href="http://yoursite.com/2021/01/18/CycleGAN%E8%AE%BA%E6%96%87%E7%B2%BE%E8%AF%BB/"/>
    <id>http://yoursite.com/2021/01/18/CycleGAN%E8%AE%BA%E6%96%87%E7%B2%BE%E8%AF%BB/</id>
    <published>2021-01-18T05:17:19.000Z</published>
    <updated>2021-01-22T14:27:58.584Z</updated>
    
    <summary type="html">
    
      &lt;h1 id=&quot;CycleGAN论文精读&quot;&gt;&lt;a href=&quot;#CycleGAN论文精读&quot; class=&quot;headerlink&quot; title=&quot;CycleGAN论文精读&quot;&gt;&lt;/a&gt;CycleGAN论文精读&lt;/h1&gt;&lt;blockquote&gt;
&lt;p&gt;学习论文：Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks&lt;/p&gt;
&lt;p&gt;论文目的：Traditional Image-to-image translation is a important topic,this paper propose a new loss to make the translation consitent in the absense of paired training datas.&lt;/p&gt;
&lt;p&gt;论文摘要: Traditional Image-to-image translation need aligned image pairs,however it’s hard to achieve.So this paper uses  Cycle Loss to solve this problem.&lt;/p&gt;
&lt;p&gt;关键词：&lt;strong&gt;CycleLoss&lt;/strong&gt;    &lt;strong&gt;GAN&lt;/strong&gt;     &lt;strong&gt;Image-to-image translation&lt;/strong&gt;&lt;/p&gt;
&lt;/blockquote&gt;
    
    </summary>
    
    
    
  </entry>
  
  <entry>
    <title>DCGAN</title>
    <link href="http://yoursite.com/2021/01/13/DCGAN/"/>
    <id>http://yoursite.com/2021/01/13/DCGAN/</id>
    <published>2021-01-13T09:48:18.000Z</published>
    <updated>2021-01-14T13:39:14.596Z</updated>
    
    <summary type="html">
    
      &lt;h1 id=&quot;DCGAN论文精读&quot;&gt;&lt;a href=&quot;#DCGAN论文精读&quot; class=&quot;headerlink&quot; title=&quot;DCGAN论文精读&quot;&gt;&lt;/a&gt;DCGAN论文精读&lt;/h1&gt;&lt;blockquote&gt;
&lt;p&gt;学习论文：UNSUPERVISED REPRESENTATION LEARNING WITH DEEP CONVOLUTIONAL GENERATIVE ADVERSARIAL NETWORKS&lt;/p&gt;
&lt;p&gt;论文目的：It brige a gap between the successs of CNN and unsupervised learning which uses GAN.&lt;/p&gt;
&lt;p&gt;论文摘要：First ,it replace the layer of Liner with CNN in GAN. Second,it demonstrate its application in general images.&lt;/p&gt;
&lt;p&gt;关键词：&lt;strong&gt;CNN&lt;/strong&gt;    &lt;strong&gt;GAN&lt;/strong&gt;&lt;/p&gt;
&lt;/blockquote&gt;
    
    </summary>
    
    
    
  </entry>
  
  <entry>
    <title>GAN论文精读</title>
    <link href="http://yoursite.com/2021/01/12/GAN%E8%AE%BA%E6%96%87%E7%B2%BE%E8%AF%BB/"/>
    <id>http://yoursite.com/2021/01/12/GAN%E8%AE%BA%E6%96%87%E7%B2%BE%E8%AF%BB/</id>
    <published>2021-01-12T07:32:10.000Z</published>
    <updated>2021-01-13T09:29:23.458Z</updated>
    
    <summary type="html">
    
      &lt;h1 id=&quot;GAN论文精读&quot;&gt;&lt;a href=&quot;#GAN论文精读&quot; class=&quot;headerlink&quot; title=&quot;GAN论文精读&quot;&gt;&lt;/a&gt;GAN论文精读&lt;/h1&gt;&lt;blockquote&gt;
&lt;p&gt;学习论文：Generative Adversarial Nets&lt;/p&gt;
&lt;p&gt;论文目的：Although the discriminative model in deep learning has make a great progress ,the generative model have had less of an impact . So this paper make a import breakthtough in genetative model .&lt;/p&gt;
&lt;p&gt;论文摘要：He propose a new framewordk called GAN(Generative Adversarial Nets).It trains two model : a generative model &lt;strong&gt;G&lt;/strong&gt; to generate the distribution of data , a discriminative model &lt;strong&gt;D&lt;/strong&gt; to estimate whether the input is fake.So the G’s goal is &lt;strong&gt;maximize the error of D&lt;/strong&gt;,the D’s goal is &lt;strong&gt;minimize the error of decision&lt;/strong&gt;.So the pefect situation is  D’s error rate is 50% because the G and the ground truth have the same distribution.&lt;/p&gt;
&lt;p&gt;关键词：&lt;strong&gt;generative model&lt;/strong&gt;&lt;/p&gt;&lt;/blockquote&gt;
    
    </summary>
    
    
    
  </entry>
  
  <entry>
    <title>PeLeeNet代码解读</title>
    <link href="http://yoursite.com/2021/01/07/PeLeeNet%E4%BB%A3%E7%A0%81%E8%A7%A3%E8%AF%BB/"/>
    <id>http://yoursite.com/2021/01/07/PeLeeNet%E4%BB%A3%E7%A0%81%E8%A7%A3%E8%AF%BB/</id>
    <published>2021-01-07T03:47:40.000Z</published>
    <updated>2021-01-08T08:25:59.468Z</updated>
    
    <summary type="html">
    
      &lt;h1 id=&quot;PeLeeNet代码解读&quot;&gt;&lt;a href=&quot;#PeLeeNet代码解读&quot; class=&quot;headerlink&quot; title=&quot;PeLeeNet代码解读&quot;&gt;&lt;/a&gt;PeLeeNet代码解读&lt;/h1&gt;&lt;blockquote&gt;
&lt;p&gt;该论文解读，也将成为母版。&lt;/p&gt;
&lt;p&gt;PeleeNet是一种基于Densenet的轻量化SSD网络变体,所以参数量更小，可以在mobile端跑。&lt;/p&gt;
&lt;p&gt;解读代码：&lt;a href=&quot;https://github.com/yxlijun/Pelee.Pytorch&quot; target=&quot;_blank&quot; rel=&quot;noopener&quot;&gt;https://github.com/yxlijun/Pelee.Pytorch&lt;/a&gt;&lt;/p&gt;
&lt;/blockquote&gt;
    
    </summary>
    
    
    
  </entry>
  
</feed>