转
人脸验证（四）--CenterLoss
有点纠结，实质上本文只是在深度学习框架下人脸识别损失函数的改进。但我还是把它归类到端到端了。出于个人原因。
转自：http://blog.csdn.net/yang_502/article/details/72792786
16年ECCV的文章《A Discriminative Feature Learning Approach for Deep Face Recognition》
code：https://github.com/ydwen/caffe-face
Motivation：
和metric learning的想法一致，希望同类样本之间紧凑，不同类样本之间分散。现有的CNN最常用的softmax损失函数来训练网络，得到的深度特征通常具有比较强的区分性，也就是比较强的类间判别力。关于softmax的类内判别力，作者在文章中给了toy example,很直观的理解。
上面给的是mnist的最后一层特征在二维空间的一个分布情况，可以看到类间是可分的，但类内存在的差距还是比较大的。对于像人脸这种复杂分布的数据，我们通常不仅希望数据在特征空间不仅是类间可分，更重要的是类内紧凑。因为同一个人的类内变化很可能会大于类间的变化，只有保持类内紧凑，我们才能对那些类内大变化的样本有一个更加鲁棒的判定结果。也就是学习一种discriminative的特征。
下图就是我们希望达到的一种效果。
另外一种通俗解释 ：
一句话：通过添加center loss 让简单的softmax 能够训练出更有内聚性的特征。
作者意图，在配合softmax适用的时候，希望使学习到的特征具有更好的泛化性和辨别能力。通过惩罚每个种类的样本和该种类样本中心的偏移，使得同一种类的样本尽量聚合在一起。
相对于triplet（Google FaceNet: A Unified Embedding for Face Recognition and Clustering：2015）和contrastive(汤晓鸥 Deep LearningFace Representation by Joint Identification-Verification：2014)来说，这个目标其实相对‘清晰’， 所以网络收敛的速度甚至比仅仅用softmax更快，而且不需要像前两者那样构造大量的训练对。
看一张图：
在左图中，我们发现一个类如果太胖，那么出现的结果就是类内距离>类间距离。（任意红点之间的距离应该小于红蓝之间的距离。）
左边时softmax一般结果，右边时centerloss结果。我们期望特征不仅可分，而且时必须差异大。如右边图。
Approach：
考虑保持softmax loss的类间判别力，提出center loss，center loss就是为了约束类内紧凑的条件。相比于传统的CNN，仅改变了原有的损失函数，易于训练和优化网络。
softmax loss :
center loss:
其中，cyi就是每类人特征的中心表示。这个就是一个类中心特征，希望在一类人的所有图像特征与类中心特征的距离总和最小。
完整的loss function：
关于center loss 的求导：
最后的loss实现可以参考code的具体实现。
Experiments：
CNN architecture
Weights in three local convolution layers are locally shared in the regions of 4 × 4, 2 × 2 and 1 × 1 respectively.
Implementation Details
Preprocessing.
使用【40】的关键点检测，进行人脸对齐与检测。人脸图像裁剪成112*96大小，像素归一化操作是通过减127.5再除128（？？没搞懂）
Training data.
文章中使用的训练数据比较多，包括有CASIA-WebFace [39], CACD2000 [4], Celebrity+ [22]。但是作者他们做了图像清理，保证和测试集的人身份不重复，最后保留17,189个人的0.7M的图像。后面的训练阶段，只用了0.49M的数据，并且对图像做了水平翻转的数据增强。
Detailed settings in CNNs.
文章比较了3中模型：we respectively train three kind of models under the supervision of softmax loss (model A), softmax loss and contrastive loss (model B), softmax loss and center loss (model C)。就是网络使用的损失函数不同。
batchsize=256，在Titan X的两块GPU上做训练。A和C的初始lr=0.1,分别在16K和24K之后缩小至0.01.文章提到Ａ在２８Ｋ次收敛大致花１４h,model B的收敛比较慢，初始lr=.01,咋２４Ｋ和３６Ｋ迭代的时候缩小学习率。Ｂ总共迭代了４２Ｋ次，花２２h.没说到Ｃ的训练时间。
Detailed settings in testing.
测试阶段，将测试图像通过ＣＮＮ提取深度特征，即最后一个ＦＣ层的输出，将原图以及原图的水平翻转图的特征图拼接表示成共同的表示，经过ＰＣＡ降维后计算ＣＯＳ距离。对于识别或者验证任务，可以是简单通过阈值划分或者是近邻判定。
Experiments on the LFW and YTF datasets
使用０．７Ｍ的训练数据，在ＬＦＷ和ＹＴＦ的pair对上做了人脸验证。
Experiments on the dataset of MegaFace Challenge
使用ＭｅｇａＦａｃｅ这个百万级别的数据库（干扰图像很多）。这个对于识别任务增大了难度。
数据库具体包括：
MegaFace datasets include gallery set and probe set. The gallery set consists of more than 1 million images from 690K different individuals, as a subset of Flickr photos [35] from Yahoo. The probe set using in this challenge are two existing databases: Facescrub [24] and FGNet [1]. Facescrub dataset is publicly available dataset, containing 100K photos of 530 unique individuals (55,742 images of males and 52,076 images of females). The possible bias can be reduced by sufficient samples in each identity. FGNet dataset is a face aging dataset, with 1002 images from 82 identities. Each identity has multiple face images at different ages (ranging from 0 to 69).
Ｒｅｆｅｒｅｎｃｅ
【４０】 ＭＴＣＮＮ
Zhang, K., Zhang, Z., Li, Z., Qiao, Y.: Joint face detection and alignment using multi-task cascaded convolutional networks. arXiv preprint arXiv:1604.02878(2016)
<script>
(function(){
function setArticleH(btnReadmore,posi){
var winH = $(window).height();
var articleBox = $("div.article_content");
var artH = articleBox.height();
if(artH > winH*posi){
articleBox.css({
'height':winH*posi+'px',
'overflow':'hidden'
})
btnReadmore.click(function(){
articleBox.removeAttr("style");
$(this).parent().remove();
})
}else{
btnReadmore.parent().remove();
}
}
var btnReadmore = $("#btn-readmore");
if(btnReadmore.length>0){
if(currentUserName){
setArticleH(btnReadmore,3);
}else{
setArticleH(btnReadmore,1.2);
}
}
})()
</script>
</article>
用基于center loss的人脸识别模型对LFW人脸数据集进行评测（c++） - 花泽
05-13 6681
接上一篇博文，这篇博文主要是进行人脸识别中的第③和第四个步骤：特征提取以及相似度计算。 center loss是2016的一篇ECCV论文中提出来的，A Discrimina... 来自： 花泽
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/oppo62258801/article/details/78034014,BlogCommendClickRateRank_1,index_1"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/oppo62258801/article/details/78034014,BlogCommendClickRateRank_1,index_1"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/oppo62258801/article/details/78034014" target="_blank" title="损失函数：Hinge Loss（max margin） - oppo62258801的博客">
<h4 class="text-truncate oneline" style="width: 610px;">
损失函数：Hinge Loss（max margin） - oppo62258801的博客 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/D/7/3/3_oppo62258801.jpg" alt="oppo62258801" class="avatar-pic">
<span class="namebox">
<span class="name">oppo62258801</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">09-19</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
1524</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/oppo62258801/article/details/78034014" target="_blank" title="损失函数：Hinge Loss（max margin） - oppo62258801的博客">
<span class="desc oneline">目录(?)[-]
Hinge Loss简介变种
Hinge Loss简介
Hinge Loss是一种目标函数（或者说损失函数）的名称，有的时候又叫做max-margin …
来自： oppo62258801的博客
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/wfei101/article/details/79773547,BlogCommendClickRateRank_2,index_2"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/wfei101/article/details/79773547,BlogCommendClickRateRank_2,index_2"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/wfei101/article/details/79773547" target="_blank" title="人脸识别：Loss函数总结 - BigCowPeking">
<h4 class="text-truncate oneline" style="width: 610px;">
<em>人脸</em>识别：Loss函数总结 - BigCowPeking </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/3/8/D/3_wfei101.jpg" alt="wfei101" class="avatar-pic">
<span class="namebox">
<span class="name">wfei101</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">03-31</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
1697</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/wfei101/article/details/79773547" target="_blank" title="人脸识别：Loss函数总结 - BigCowPeking">
<span class="desc oneline">基于深度学习的人脸识别发展，从deepid开始，到今年（或者说去年），已经基本趋于成熟。凡是基于识别的，总是离不开三个东西：数据，网络，以及loss。数据方面，目前的公开数据集中有主打类别数的MS_c...</span>
</a>
<span class="blog_title_box oneline"><a target="_blank" href="https://blog.csdn.net/wfei101">来自： <span class="blog_title"> BigCowPeking</span></a></span>
</p>
</div>
</div>
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/tinyzhao/article/details/53236191,BlogCommendClickRateRank_3,index_3"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/tinyzhao/article/details/53236191,BlogCommendClickRateRank_3,index_3"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/tinyzhao/article/details/53236191" target="_blank" title="人脸检测：MTCNN - flyingzhao">
<h4 class="text-truncate oneline" style="width: 608px;">
<em>人脸</em>检测：MTCNN - flyingzhao </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/7/3/0/3_tinyzhao.jpg" alt="tinyzhao" class="avatar-pic">
<span class="namebox">
<span class="name">tinyzhao</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">11-19</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
3.2万</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/tinyzhao/article/details/53236191" target="_blank" title="人脸检测：MTCNN - flyingzhao">
<span class="desc oneline">《Joint Face Detection and Alignment using Multi-task Cascaded Convolutional Networks》论文解读。本文来自于中国科学院...</span>
</a>
<span class="blog_title_box oneline"><a target="_blank" href="https://blog.csdn.net/tinyzhao">来自： <span class="blog_title"> flyingzhao</span></a></span>
</p>
</div>
</div>
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/angus_17/article/details/78894180,BlogCommendClickRateRank_4,index_4"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/angus_17/article/details/78894180,BlogCommendClickRateRank_4,index_4"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/angus_17/article/details/78894180" target="_blank" title="人脸识别领域中阈值的计算 - 漂泊小柒的专栏">
<h4 class="text-truncate oneline" style="width: 610px;">
<em>人脸</em>识别领域中阈值的计算 - 漂泊小柒的专栏 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/D/8/C/3_angus_17.jpg" alt="angus_17" class="avatar-pic">
<span class="namebox">
<span class="name">angus_17</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">12-25</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
1033</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/angus_17/article/details/78894180" target="_blank" title="人脸识别领域中阈值的计算 - 漂泊小柒的专栏">
<span class="desc oneline">举一个实际案例，机场安检前台，用的1:1核身比对，正样本是正常安检用户，负样本是持他人护照的客户（攻击！）
会这么思考问题：
a.这个场景下，正样本远远多于负样本（攻击极少），比如十万比一；
b…
来自： 漂泊小柒的专栏
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/wfei101/article/details/79252067,BlogCommendClickRateRank_5,index_5"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/wfei101/article/details/79252067,BlogCommendClickRateRank_5,index_5"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/wfei101/article/details/79252067" target="_blank" title="人脸识别：损失函数之softmax loss和cross entropy Loss - BigCowPeking">
<h4 class="text-truncate oneline" style="width: 610px;">
<em>人脸</em>识别：损失函数之softmax loss和cross entropy Loss - BigCowPeking </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/3/8/D/3_wfei101.jpg" alt="wfei101" class="avatar-pic">
<span class="namebox">
<span class="name">wfei101</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">02-04</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
1404</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/wfei101/article/details/79252067" target="_blank" title="人脸识别：损失函数之softmax loss和cross entropy Loss - BigCowPeking">
<span class="desc oneline">转载blog：http://blog.csdn.net/u014380165/article/details/77284921
我们知道卷积神经网络（CNN）在图像领域的应用已经非常广泛了，一般…
来自： BigCowPeking
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/qq_14845119/article/details/53308996,BlogCommendClickRateRank_6,index_6"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/qq_14845119/article/details/53308996,BlogCommendClickRateRank_6,index_6"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/qq_14845119/article/details/53308996" target="_blank" title="人脸识别之caffe-face - 年轻即出发，">
<h4 class="text-truncate oneline" style="width: 608px;">
<em>人脸</em>识别之caffe-face - 年轻即出发， </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/5/B/9/3_qq_14845119.jpg" alt="qq_14845119" class="avatar-pic">
<span class="namebox">
<span class="name">qq_14845119</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">11-23</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
1.9万</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/qq_14845119/article/details/53308996" target="_blank" title="人脸识别之caffe-face - 年轻即出发，">
<span class="desc oneline">该论文为2016年的一篇ECCV，ADiscriminative Feature Learning Approach for Deep Face Recognition。还是深圳先进院乔宇老师组的一篇...</span>
</a>
<span class="blog_title_box oneline"><a target="_blank" href="https://blog.csdn.net/qq_14845119">来自： <span class="blog_title"> 年轻即出发，</span></a></span>
</p>
</div>
</div>
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/dreamer_on_air/article/details/78521071,BlogCommendClickRateRank_7,index_7"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/dreamer_on_air/article/details/78521071,BlogCommendClickRateRank_7,index_7"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/dreamer_on_air/article/details/78521071" target="_blank" title="人脸识别：coco loss-Rethinking Feature Discrimination and Polymerization for Large-scale Recognition - 迷途中一个小书童">
<h4 class="text-truncate oneline" style="width: 610px;">
<em>人脸</em>识别：coco loss-Rethinking Feature Discrimination and Polymerization for Large-scale Recognition - 迷途中一个小书童 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/8/A/D/3_dreamer_on_air.jpg" alt="dreamer_on_air" class="avatar-pic">
<span class="namebox">
<span class="name">dreamer_on_air</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">11-13</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
1443</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/dreamer_on_air/article/details/78521071" target="_blank" title="人脸识别：coco loss-Rethinking Feature Discrimination and Polymerization for Large-scale Recognition - 迷途中一个小书童">
<span class="desc oneline">nips的一篇做分类和识别的工作，其中在人脸识别任务上也做了实验，Rethinking Feature Discrimination and Polymerization for Large-scal...</span>
</a>
<span class="blog_title_box oneline"><a target="_blank" href="https://blog.csdn.net/dreamer_on_air">来自： <span class="blog_title"> 迷途中一个小书童</span></a></span>
</p>
</div>
</div>
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/wfei101/article/details/79773493,BlogCommendClickRateRank_8,index_8"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/wfei101/article/details/79773493,BlogCommendClickRateRank_8,index_8"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/wfei101/article/details/79773493" target="_blank" title="目标检测之Loss：Center Loss梯度更新 - BigCowPeking">
<h4 class="text-truncate oneline" style="width: 618px;">
目标检测之Loss：Center Loss梯度更新 - BigCowPeking </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/3/8/D/3_wfei101.jpg" alt="wfei101" class="avatar-pic">
<span class="namebox">
<span class="name">wfei101</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">03-31</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
442</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/wfei101/article/details/79773493" target="_blank" title="目标检测之Loss：Center Loss梯度更新 - BigCowPeking">
<span class="desc oneline">转载：https://blog.csdn.net/u014380165/article/details/76946339最近几年网络效果的提升除了改变网络结构外，还有一群人在研究损失层的改进，这篇博文...</span>
</a>
<span class="blog_title_box oneline"><a target="_blank" href="https://blog.csdn.net/wfei101">来自： <span class="blog_title"> BigCowPeking</span></a></span>
</p>
</div>
</div>
<div class="recommend-item-box type_hot_word">
<div class="content clearfix oneline" style="width: 702px;">
<h5 class="float-left">文章热词</h5>
<div class="word float-left">
<span>
<a href="https://edu.csdn.net/courses/o5329_s5330_k " target="_blank">
机器学习 </a></span>
<span>
<a href="https://edu.csdn.net/courses/o5329_s5330_k " target="_blank">
机器学习课程 </a></span>
<span>
<a href="https://edu.csdn.net/courses/o5329_s5330_k " target="_blank">
机器学习教程 </a></span>
<span>
<a href="https://edu.csdn.net/combos/o5329_s5331_l0_t " target="_blank">
深度学习视频教程 </a></span>
<span>
<a href="https://edu.csdn.net/combos/o5329_s5331_l0_t " target="_blank">
深度学习学习 </a></span>
</div>
</div>
<div class="content clearfix oneline" style="width: 702px;">
<h5 class="float-left">相关热词</h5>
<div class="float-left">
<span>
<a href="https://blog.csdn.net/qq_17278169/article/details/79510209" target="_blank">
人脸</a>
</span>
<span>
<a href="https://blog.csdn.net/Trent1985/article/details/79760854" target="_blank">
人脸檄文人脸</a>
</span>
<span>
<a href="https://blog.csdn.net/c2a2o2/article/details/72648876" target="_blank">
按照人脸生成人脸</a>
</span>
<span>
<a href="https://blog.csdn.net/wfei101/article/details/80489119" target="_blank">
人脸模型检测人脸</a>
</span>
<span>
<a href="https://blog.csdn.net/as_far_as/article/details/81391867" target="_blank">
百度人脸V3人脸</a>
</span>
</div>
</div>
</div>
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/Fire_Light_/article/details/79598497,BlogCommendClickRateRank_9,index_9"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/Fire_Light_/article/details/79598497,BlogCommendClickRateRank_9,index_9"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/Fire_Light_/article/details/79598497" target="_blank" title="人脸识别系列（十二）：Center Loss - Fire_Light_的博客">
<h4 class="text-truncate oneline" style="width: 610px;">
<em>人脸</em>识别系列（十二）：Center Loss - Fire_Light_的博客 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/8/2/5/3_fire_light_.jpg" alt="Fire_Light_" class="avatar-pic">
<span class="namebox">
<span class="name">Fire_Light_</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">03-18</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
1378</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/Fire_Light_/article/details/79598497" target="_blank" title="人脸识别系列（十二）：Center Loss - Fire_Light_的博客">
<span class="desc oneline">论文链接：A Discriminative Feature Learning Approach for Deep Face Recognition（需翻墙）
代码链接:https://github…
来自： Fire_Light_的博客
<div class="recommend-item-box blog-expert-recommend-box" style="display: block;">
<div class="d-flex">
<div class="blog-expert-recommend">
<div class="blog-expert">
<div class="blog-expert-flexbox" data-track-view="{"mod":"popu_709","con":"https://blog.csdn.net/app_12062011/article/details/78763892"}"><div class="blog-expert-item"><div class="blog-expert-info-box"><div class="blog-expert-img-box" data-track-click="{"mod":"popu_709","con":"https://blog.csdn.net/app_12062011/article/details/78763892"}"><a href="https://blog.csdn.net/App_12062011" target="_blank"><img src="https://avatar.csdn.net/9/0/8/3_app_12062011.jpg" alt="EasonApp"><svg class="icon" aria-hidden="true"><use xlink:href="#csdnc-blogexpert"></use></svg></a></div><div class="info"><span data-track-click="{"mod":"popu_709","con":"https://blog.csdn.net/app_12062011/article/details/78763892"}"><a href="https://blog.csdn.net/App_12062011" target="_blank"><h5 class="oneline">EasonApp</h5></a></span> <p><span data-track-click="{"mod":"popu_710","con":"https://blog.csdn.net/app_12062011/article/details/78763892"}"><span class="blog-expert-button-follow btn-red-follow" data-name="App_12062011" data-nick="EasonApp">关注</span></span><span class="article-num"> 590篇文章</span></p></div></div></div><div class="blog-expert-item"><div class="blog-expert-info-box"><div class="blog-expert-img-box" data-track-click="{"mod":"popu_709","con":"https://blog.csdn.net/app_12062011/article/details/78763892"}"><a href="https://blog.csdn.net/zoroduyu" target="_blank"><img src="https://avatar.csdn.net/3/1/8/3_zoroduyu.jpg" alt="Forward_duyu"></a></div><div class="info"><span data-track-click="{"mod":"popu_709","con":"https://blog.csdn.net/app_12062011/article/details/78763892"}"><a href="https://blog.csdn.net/zoroduyu" target="_blank"><h5 class="oneline">Forward_duyu</h5></a></span> <p><span data-track-click="{"mod":"popu_710","con":"https://blog.csdn.net/app_12062011/article/details/78763892"}"><span class="blog-expert-button-follow btn-red-follow" data-name="zoroduyu" data-nick="Forward_duyu">关注</span></span><span class="article-num"> 15篇文章</span></p></div></div></div><div class="blog-expert-item"><div class="blog-expert-info-box"><div class="blog-expert-img-box" data-track-click="{"mod":"popu_709","con":"https://blog.csdn.net/app_12062011/article/details/78763892"}"><a href="https://blog.csdn.net/Muzythoof" target="_blank"><img src="https://avatar.csdn.net/1/E/9/3_muzythoof.jpg" alt="Muzythoof"></a></div><div class="info"><span data-track-click="{"mod":"popu_709","con":"https://blog.csdn.net/app_12062011/article/details/78763892"}"><a href="https://blog.csdn.net/Muzythoof" target="_blank"><h5 class="oneline">Muzythoof</h5></a></span> <p><span data-track-click="{"mod":"popu_710","con":"https://blog.csdn.net/app_12062011/article/details/78763892"}"><span class="blog-expert-button-follow btn-red-follow" data-name="Muzythoof" data-nick="Muzythoof">关注</span></span><span class="article-num"> 4篇文章</span></p></div></div></div></div>
</div>
</div>
<div class="blog-expert-load-new" data-index="3" style="display: block;">
<svg version="1.1" viewBox="0 0 200 200" style="enable-background:new 0 0 200 200;" xml:space="preserve">
<path d="M142.5,57.5c-1.3-1.2-2.1-3.3-2.1-5.4c0-4.2,3.3-7.5,7.5-7.5h19.6c4.2,0,7.5,3.3,7.5,7.5s-3.3,7.5-7.5,7.5h-4.2
C170.8,71.2,175,85,175,100c0,41.7-33.8,75-75,75c-4.2,0-7.5-3.3-7.5-7.5c0-4.2,3.3-7.5,7.5-7.5c33.3,0,60-27.1,60-60
C160,83.7,153.8,68.7,142.5,57.5L142.5,57.5z M57.5,142.5c1.2,1.2,2.1,3.3,2.1,5.4c0,4.2-3.3,7.5-7.5,7.5H32.5
c-4.2,0-7.5-3.3-7.5-7.5c0-4.2,3.3-7.5,7.5-7.5h4.2C29.2,128.8,25,115,25,100c0-41.7,33.7-75,75-75c4.2,0,7.5,3.3,7.5,7.5
S104.2,40,100,40c-33.3,0-60,27.1-60,60C40,116.2,46.2,131.2,57.5,142.5z"></path>
</svg>
<span>换一批</span>
</div>
</div>
</div>
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/gdmmzmj/article/details/76100051,BlogCommendClickRateRank_10,index_10"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/gdmmzmj/article/details/76100051,BlogCommendClickRateRank_10,index_10"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/gdmmzmj/article/details/76100051" target="_blank" title="centerloss 理解 - 专注于机器学习，深度学习，人脸识别领域。">
<h4 class="text-truncate oneline" style="width: 610px;">
<em>centerloss</em> 理解 - 专注于机器学习，深度学习，<em>人脸</em>识别领域。 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/6/3/B/3_gdmmzmj.jpg" alt="gdmmzmj" class="avatar-pic">
<span class="namebox">
<span class="name">gdmmzmj</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">07-26</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
4536</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/gdmmzmj/article/details/76100051" target="_blank" title="centerloss 理解 - 专注于机器学习，深度学习，人脸识别领域。">
<span class="desc oneline">今天看了一下centerloss 源码 ，觉得还是挺值得借鉴的，所以分享一下。
centerloss主要在caffe的基础上加了三项。一. caffe.proto1.加proto 格式 的 ID…
来自： 专注于机器学习，深度学习，人脸识别领域。
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/linolzhang/article/details/71075756,BlogCommendClickRateRank_11,index_11"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/linolzhang/article/details/71075756,BlogCommendClickRateRank_11,index_11"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/linolzhang/article/details/71075756" target="_blank" title="基于深度学习的Person Re-ID（综述） - 跟随技术的脚步-linolzhang的专栏">
<h4 class="text-truncate oneline" style="width: 608px;">
基于深度学习的Person Re-ID（综述） - 跟随技术的脚步-linolzhang的专栏 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/F/A/6/3_linolzhang.jpg" alt="linolzhang" class="avatar-pic">
<span class="namebox">
<span class="name">linolzhang</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">05-01</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
1.2万</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/linolzhang/article/details/71075756" target="_blank" title="基于深度学习的Person Re-ID（综述） - 跟随技术的脚步-linolzhang的专栏">
<span class="desc oneline">一. 问题的提出
Person Re-ID 全称是 Person Re-Identification，又称为
行人重检测 or 行人再识别，直观上可以通过两种思路进行比对，一种是 通过…
来自： 跟随技术的脚步-linolzhang的专栏
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/wfei101/article/details/79251987,BlogCommendClickRateRank_12,index_12"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/wfei101/article/details/79251987,BlogCommendClickRateRank_12,index_12"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/wfei101/article/details/79251987" target="_blank" title="人脸识别：损失函数之softmax函数 - BigCowPeking">
<h4 class="text-truncate oneline" style="width: 610px;">
<em>人脸</em>识别：损失函数之softmax函数 - BigCowPeking </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/3/8/D/3_wfei101.jpg" alt="wfei101" class="avatar-pic">
<span class="namebox">
<span class="name">wfei101</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">02-04</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
1035</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/wfei101/article/details/79251987" target="_blank" title="人脸识别：损失函数之softmax函数 - BigCowPeking">
<span class="desc oneline">Loss Function
softmax_loss的计算包含2步：
（1）计算softmax归一化概率
（2）计算损失
这里以batchsize=1的2分类为…
来自： BigCowPeking
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/u011747846/article/details/80602786,BlogCommendClickRateRank_13,index_13"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/u011747846/article/details/80602786,BlogCommendClickRateRank_13,index_13"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/u011747846/article/details/80602786" target="_blank" title="Center-Loss - 当凌绝顶,俯瞰天下">
<h4 class="text-truncate oneline" style="width: 610px;">
Center-Loss - 当凌绝顶,俯瞰天下 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/E/E/D/3_u011747846.jpg" alt="u011747846" class="avatar-pic">
<span class="namebox">
<span class="name">u011747846</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">06-07</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
1020</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/u011747846/article/details/80602786" target="_blank" title="Center-Loss - 当凌绝顶,俯瞰天下">
<span class="desc oneline">摘要
这篇博客是对论文A Discriminative Feature Learning Approach for Deep Face Recognition的总结。这篇论文中，作者提出了一种新的…
来自： 当凌绝顶,俯瞰天下
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/u011808673/article/details/80491361,BlogCommendClickRateRank_14,index_14"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/u011808673/article/details/80491361,BlogCommendClickRateRank_14,index_14"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/u011808673/article/details/80491361" target="_blank" title="人脸识别-损失函数改进方法之L-softmax loss center-loss A-softmax loss focal loss - Nine days">
<h4 class="text-truncate oneline" style="width: 618px;">
<em>人脸</em>识别-损失函数改进方法之L-softmax loss center-loss A-softmax loss focal loss - Nine days </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/6/7/3/3_u011808673.jpg" alt="u011808673" class="avatar-pic">
<span class="namebox">
<span class="name">u011808673</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">05-29</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
708</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/u011808673/article/details/80491361" target="_blank" title="人脸识别-损失函数改进方法之L-softmax loss center-loss A-softmax loss focal loss - Nine days">
<span class="desc oneline">
而ArcFac中的loss与其他loss不同之处在于下图：
与CosFace的解释相同，只是ArcFace…
来自： Nine days
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/intflojx/article/details/82455560,BlogCommendClickRateRank_15,index_15"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/intflojx/article/details/82455560,BlogCommendClickRateRank_15,index_15"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/intflojx/article/details/82455560" target="_blank" title="COCO loss （人脸识别损失函数） - intflojx的博客">
<h4 class="text-truncate oneline" style="width: 618px;">
COCO loss （<em>人脸</em>识别损失函数） - intflojx的博客 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/6/5/B/3_intflojx.jpg" alt="intflojx" class="avatar-pic">
<span class="namebox">
<span class="name">intflojx</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">09-06</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
129</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/intflojx/article/details/82455560" target="_blank" title="COCO loss （人脸识别损失函数） - intflojx的博客">
<span class="desc oneline"> 2017年nips的一篇做分类和识别的工作，其中在人脸识别任务上也做了实验，Rethinking Feature Discrimination and Polymerization for Larg...</span>
</a>
<span class="blog_title_box oneline"><a target="_blank" href="https://blog.csdn.net/intflojx">来自： <span class="blog_title"> intflojx的博客</span></a></span>
</p>
</div>
</div>
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/qq_41185868/article/details/82929877,BlogCommendClickRateRank_16,index_16"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/qq_41185868/article/details/82929877,BlogCommendClickRateRank_16,index_16"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/qq_41185868/article/details/82929877" target="_blank" title="CV之FRec之Loss：人脸识别中常用的几种损失简介、使用方法之详细攻略 - 一个处女座的程序猿">
<h4 class="text-truncate oneline" style="width: 626px;">
CV之FRec之Loss：<em>人脸</em>识别中常用的几种损失简介、使用方法之详细攻略 - 一个处女座的程序猿 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/B/4/2/3_qq_41185868.jpg" alt="qq_41185868" class="avatar-pic">
<span class="namebox">
<span class="name">qq_41185868</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">10-03</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
77</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/qq_41185868/article/details/82929877" target="_blank" title="CV之FRec之Loss：人脸识别中常用的几种损失简介、使用方法之详细攻略 - 一个处女座的程序猿">
<span class="desc oneline">CV之FRec之Loss：人脸识别中常用的几种损失简介、使用方法之详细攻略
T1、Triplet Loss
《FaceNet: A Unified Embedding for Face Re…
来自： 一个处女座的程序猿
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/Iriving_shu/article/details/77822201,BlogCommendClickRateRank_17,index_17"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/Iriving_shu/article/details/77822201,BlogCommendClickRateRank_17,index_17"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/Iriving_shu/article/details/77822201" target="_blank" title="人脸识别系列-Centerloss论文阅读笔记 - Iriving_shu的博客">
<h4 class="text-truncate oneline" style="width: 618px;">
<em>人脸</em>识别系列-<em>Centerloss</em>论文阅读笔记 - Iriving_shu的博客 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/B/3/9/3_iriving_shu.jpg" alt="Iriving_shu" class="avatar-pic">
<span class="namebox">
<span class="name">Iriving_shu</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">09-03</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
876</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/Iriving_shu/article/details/77822201" target="_blank" title="人脸识别系列-Centerloss论文阅读笔记 - Iriving_shu的博客">
<span class="desc oneline">好记心不如烂笔头，以此记录自己的学习历程。
这篇论文 Centerloss来自Eccv2016。本文主要讲解其核心原理以及源码和实验。
论文地址：https://ydwen.github.io/p…
来自： Iriving_shu的博客
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/u014380165/article/details/76946339,BlogCommendClickRateRank_18,index_18"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/u014380165/article/details/76946339,BlogCommendClickRateRank_18,index_18"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/u014380165/article/details/76946339" target="_blank" title="损失函数改进之Center Loss - AI之路">
<h4 class="text-truncate oneline" style="width: 608px;">
损失函数改进之Center Loss - AI之路 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/D/6/1/3_u014380165.jpg" alt="u014380165" class="avatar-pic">
<span class="namebox">
<span class="name">u014380165</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">08-09</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
1.8万</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/u014380165/article/details/76946339" target="_blank" title="损失函数改进之Center Loss - AI之路">
<span class="desc oneline">最近几年网络效果的提升除了改变网络结构外，还有一群人在研究损失层的改进，这篇博文要介绍的就是较为新颖的center loss。center loss来自ECCV2016的一篇论文：A Discrimi...</span>
</a>
<span class="blog_title_box oneline"><a target="_blank" href="https://blog.csdn.net/u014380165">来自： <span class="blog_title"> AI之路</span></a></span>
</p>
</div>
</div>
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/leadai/article/details/78876511,BlogCommendClickRateRank_19,index_19"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/leadai/article/details/78876511,BlogCommendClickRateRank_19,index_19"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/leadai/article/details/78876511" target="_blank" title="深度学习中的损失函数总结以及Center Loss函数笔记 - leadai的博客">
<h4 class="text-truncate oneline" style="width: 610px;">
深度学习中的损失函数总结以及Center Loss函数笔记 - leadai的博客 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/5/0/5/3_leadai.jpg" alt="leadai" class="avatar-pic">
<span class="namebox">
<span class="name">leadai</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">12-22</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
4961</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/leadai/article/details/78876511" target="_blank" title="深度学习中的损失函数总结以及Center Loss函数笔记 - leadai的博客">
<span class="desc oneline">北京 上海巡回站 | NVIDIA DLI深度学习培训2018年1月26/1月12日NVIDIA 深度学习学院 带你快速进入火热的DL领域阅读全文 ...</span>
</a>
<span class="blog_title_box oneline"><a target="_blank" href="https://blog.csdn.net/leadai">来自： <span class="blog_title"> leadai的博客</span></a></span>
</p>
</div>
</div>
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/marleylee/article/details/81212659,BlogCommendClickRateRank_20,index_20"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/marleylee/article/details/81212659,BlogCommendClickRateRank_20,index_20"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/marleylee/article/details/81212659" target="_blank" title="人脸识别之人脸验证（四）--CenterLoss - marleylee的博客">
<h4 class="text-truncate oneline" style="width: 626px;">
<em>人脸</em>识别之<em>人脸</em><em>验证</em>（四）--<em>CenterLoss</em> - marleylee的博客 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/B/B/9/3_marleylee.jpg" alt="marleylee" class="avatar-pic">
<span class="namebox">
<span class="name">marleylee</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">07-26</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
37</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/marleylee/article/details/81212659" target="_blank" title="人脸识别之人脸验证（四）--CenterLoss - marleylee的博客">
<span class="desc oneline">实质上本文只是在深度学习框架下人脸识别损失函数的改进。但这里还是把它归类到端到端了。
转自：http://blog.csdn.net/yang_502/article/details/7279…
来自： marleylee的博客
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/MrCharles/article/details/80096207,BlogCommendClickRateRank_21,index_21"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/MrCharles/article/details/80096207,BlogCommendClickRateRank_21,index_21"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/MrCharles/article/details/80096207" target="_blank" title="人脸识别的LOSS（上） - MrCharles--Code is Poetry">
<h4 class="text-truncate oneline" style="width: 618px;">
<em>人脸</em>识别的LOSS（上） - MrCharles--Code is Poetry </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/5/D/C/3_mrcharles.jpg" alt="MrCharles" class="avatar-pic">
<span class="namebox">
<span class="name">MrCharles</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">04-26</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
262</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/MrCharles/article/details/80096207" target="_blank" title="人脸识别的LOSS（上） - MrCharles--Code is Poetry">
<span class="desc oneline">超多分类的Softmax
2014年CVPR两篇超多分类的人脸识别论文：DeepFace和DeepID
Taigman Y, Yang M, Ranzato M A, et al.
De…
来自： MrCharles–Code is Poetry
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/App_12062011/article/details/78510595,BlogCommendClickRateRank_22,index_22"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/App_12062011/article/details/78510595,BlogCommendClickRateRank_22,index_22"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/App_12062011/article/details/78510595" target="_blank" title="人脸验证（二）--DeepID - 工作笔记">
<h4 class="text-truncate oneline" style="width: 610px;">
<em>人脸</em><em>验证</em>（二）--DeepID - 工作笔记 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/9/0/8/3_app_12062011.jpg" alt="App_12062011" class="avatar-pic">
<span class="namebox">
<span class="name">App_12062011</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">11-12</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
1534</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/App_12062011/article/details/78510595" target="_blank" title="人脸验证（二）--DeepID - 工作笔记">
<span class="desc oneline">转自：http://blog.csdn.net/stdcoutzyx/article/details/42091205
- 问题引入及算法流程
DeepID所应用的领域是人脸识别的子领域——人脸…
来自： 工作笔记
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/intflojx/article/details/82377939,BlogCommendClickRateRank_23,index_23"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/intflojx/article/details/82377939,BlogCommendClickRateRank_23,index_23"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/intflojx/article/details/82377939" target="_blank" title="人脸识别之数据、网络结构、损失函数 - intflojx的博客">
<h4 class="text-truncate oneline" style="width: 618px;">
<em>人脸</em>识别之数据、网络结构、损失函数 - intflojx的博客 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/6/5/B/3_intflojx.jpg" alt="intflojx" class="avatar-pic">
<span class="namebox">
<span class="name">intflojx</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">09-04</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
138</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/intflojx/article/details/82377939" target="_blank" title="人脸识别之数据、网络结构、损失函数 - intflojx的博客">
<span class="desc oneline">基于深度学习的人脸识别发展，已经基本趋于成熟，业界比较成熟的人脸1：N应用总结如下：
https://blog.csdn.net/intflojx/article/details/81278330
…
来自： intflojx的博客
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/weixin_41278720/article/details/79848433,BlogCommendClickRateRank_24,index_24"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/weixin_41278720/article/details/79848433,BlogCommendClickRateRank_24,index_24"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/weixin_41278720/article/details/79848433" target="_blank" title="新型损失函数LMCL：可显著增强人脸识别模型的判别能力 - spectre">
<h4 class="text-truncate oneline" style="width: 618px;">
新型损失函数LMCL：可显著增强<em>人脸</em>识别模型的判别能力 - spectre </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/2/2/1/3_weixin_41278720.jpg" alt="weixin_41278720" class="avatar-pic">
<span class="namebox">
<span class="name">weixin_41278720</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">04-08</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
248</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/weixin_41278720/article/details/79848433" target="_blank" title="新型损失函数LMCL：可显著增强人脸识别模型的判别能力 - spectre">
<span class="desc oneline">原标题:CVPR 2018 | 腾讯AI Lab提出新型损失函数LMCL：可显著增强人脸识别模型的判别能力 选自arXiv 机器之心编译 深度卷积神经网络 (CNN) 已经推动人脸识别实现了革命...</span>
</a>
<span class="blog_title_box oneline"><a target="_blank" href="https://blog.csdn.net/weixin_41278720">来自： <span class="blog_title"> spectre</span></a></span>
</p>
</div>
</div>
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/sinat_24143931/article/details/79033521,BlogCommendClickRateRank_25,index_25"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/sinat_24143931/article/details/79033521,BlogCommendClickRateRank_25,index_25"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/sinat_24143931/article/details/79033521" target="_blank" title="损失函数改进方法之A-softmax loss - 程序员深度学习">
<h4 class="text-truncate oneline" style="width: 610px;">
损失函数改进方法之A-softmax loss - 程序员深度学习 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/2/7/F/3_sinat_24143931.jpg" alt="sinat_24143931" class="avatar-pic">
<span class="namebox">
<span class="name">sinat_24143931</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">01-11</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
1301</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/sinat_24143931/article/details/79033521" target="_blank" title="损失函数改进方法之A-softmax loss - 程序员深度学习">
<span class="desc oneline">深度学习交流QQ群：116270156
论文：SphereFace: Deep Hypersphere Embedding for Face Recognition
论文链接：https…
来自： 程序员深度学习
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/He_is_all/article/details/60331997,BlogCommendClickRateRank_26,index_26"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/He_is_all/article/details/60331997,BlogCommendClickRateRank_26,index_26"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/He_is_all/article/details/60331997" target="_blank" title="论文笔记之---Joint Detection and Identification Feature Learning for Person Search - He_is_all的博客">
<h4 class="text-truncate oneline" style="width: 610px;">
论文笔记之---Joint Detection and Identification Feature Learning for Person Search - He_is_all的博客 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/1/3/F/3_he_is_all.jpg" alt="He_is_all" class="avatar-pic">
<span class="namebox">
<span class="name">He_is_all</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">03-04</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
5360</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/He_is_all/article/details/60331997" target="_blank" title="论文笔记之---Joint Detection and Identification Feature Learning for Person Search - He_is_all的博客">
<span class="desc oneline">Abstract：
现有的行人再识别方法中所使用的标准和方法主要关注的是经过裁剪的行人照片，这与现实生活场景中的图片有所不同。本文为了缩小两者的差距，提出了一种行人搜索的新框架，将行人检测和行人再识别…
来自： He_is_all的博客
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/zoroduyu/article/details/79053664,BlogCommendClickRateRank_27,index_27"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/zoroduyu/article/details/79053664,BlogCommendClickRateRank_27,index_27"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/zoroduyu/article/details/79053664" target="_blank" title="人脸识别（四）考生考试人脸验证的实现。 - zoroduyu的博客">
<h4 class="text-truncate oneline" style="width: 610px;">
<em>人脸</em>识别（四）考生考试<em>人脸</em><em>验证</em>的实现。 - zoroduyu的博客 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/3/1/8/3_zoroduyu.jpg" alt="zoroduyu" class="avatar-pic">
<span class="namebox">
<span class="name">zoroduyu</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">01-13</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
2239</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/zoroduyu/article/details/79053664" target="_blank" title="人脸识别（四）考生考试人脸验证的实现。 - zoroduyu的博客">
<span class="desc oneline">需求：
考生需要在考试前将自己的人脸录入到系统当中。（拍照录入或者直接导入，方法二选一）
考生在考试时，需要进行人脸识别，通过人脸进行身份验证，验证成功后，登录成功。
上周的博客中我讲到了如何…
来自： zoroduyu的博客
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/Muzythoof/article/details/79976584,BlogCommendClickRateRank_28,index_28"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/Muzythoof/article/details/79976584,BlogCommendClickRateRank_28,index_28"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/Muzythoof/article/details/79976584" target="_blank" title="Caffe(caffe-fast-rcnn）配置Center Loss(其他网络层类似方法） - Muzythoof的博客">
<h4 class="text-truncate oneline" style="width: 618px;">
Caffe(caffe-fast-rcnn）配置Center Loss(其他网络层类似方法） - Muzythoof的博客 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/1/E/9/3_muzythoof.jpg" alt="Muzythoof" class="avatar-pic">
<span class="namebox">
<span class="name">Muzythoof</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">04-17</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
260</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/Muzythoof/article/details/79976584" target="_blank" title="Caffe(caffe-fast-rcnn）配置Center Loss(其他网络层类似方法） - Muzythoof的博客">
<span class="desc oneline">一、首先是配置CenterLoss原版的caffe是没有Centerloss这一层的。所以需要手动添加。参考博客：http://blog.csdn.net/samylee/article/detail...</span>
</a>
<span class="blog_title_box oneline"><a target="_blank" href="https://blog.csdn.net/Muzythoof">来自： <span class="blog_title"> Muzythoof的博客</span></a></span>
</p>
</div>
</div>
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/zoroduyu/article/details/78994826,BlogCommendClickRateRank_29,index_29"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/zoroduyu/article/details/78994826,BlogCommendClickRateRank_29,index_29"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/zoroduyu/article/details/78994826" target="_blank" title="人脸识别（三）模拟业务场景：考生考试人脸验证的实现。 - zoroduyu的博客">
<h4 class="text-truncate oneline" style="width: 610px;">
<em>人脸</em>识别（三）模拟业务场景：考生考试<em>人脸</em><em>验证</em>的实现。 - zoroduyu的博客 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/3/1/8/3_zoroduyu.jpg" alt="zoroduyu" class="avatar-pic">
<span class="namebox">
<span class="name">zoroduyu</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">01-07</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
1085</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/zoroduyu/article/details/78994826" target="_blank" title="人脸识别（三）模拟业务场景：考生考试人脸验证的实现。 - zoroduyu的博客">
<span class="desc oneline">在上篇博客的结尾处，我提到了这么两个需求：
我在之前博客中讲到的face++的api可以用来实现这么这两个功能，今天这篇博客我们讲来讲一讲具体的实现。
先来看第一个需求：
考生需要在考试前…
来自： zoroduyu的博客
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/Yan_Joy/article/details/79104993,BlogCommendClickRateRank_30,index_30"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/Yan_Joy/article/details/79104993,BlogCommendClickRateRank_30,index_30"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/Yan_Joy/article/details/79104993" target="_blank" title="center loss 论文学习 - Yan_Joy的博客">
<h4 class="text-truncate oneline" style="width: 618px;">
center loss 论文学习 - Yan_Joy的博客 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/5/6/E/3_yan_joy.jpg" alt="Yan_Joy" class="avatar-pic">
<span class="namebox">
<span class="name">Yan_Joy</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">02-27</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
347</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/Yan_Joy/article/details/79104993" target="_blank" title="center loss 论文学习 - Yan_Joy的博客">
<span class="desc oneline">center loss框架从网络的的框架来看，center loss的主要工作是下图中的“Discriminative Features”。
普通的网络框架，在反向传播的过程中，根据类别标签，会将不…
来自： Yan_Joy的博客
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/tinyzhao/article/details/53002441,BlogCommendClickRateRank_31,index_31"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/tinyzhao/article/details/53002441,BlogCommendClickRateRank_31,index_31"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/tinyzhao/article/details/53002441" target="_blank" title="人脸验证：DeepID - flyingzhao">
<h4 class="text-truncate oneline" style="width: 610px;">
<em>人脸</em><em>验证</em>：DeepID - flyingzhao </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/7/3/0/3_tinyzhao.jpg" alt="tinyzhao" class="avatar-pic">
<span class="namebox">
<span class="name">tinyzhao</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">11-01</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
6832</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/tinyzhao/article/details/53002441" target="_blank" title="人脸验证：DeepID - flyingzhao">
<span class="desc oneline">《Deep Learning Face Representation from Predicting 10000 Classes》论文解读本文来自香港中文大学汤晓鸥课题组。人脸验证（Face Veri...</span>
</a>
<span class="blog_title_box oneline"><a target="_blank" href="https://blog.csdn.net/tinyzhao">来自： <span class="blog_title"> flyingzhao</span></a></span>
</p>
</div>
</div>
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/Lucifer_zzq/article/details/81236174,BlogCommendClickRateRank_32,index_32"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/Lucifer_zzq/article/details/81236174,BlogCommendClickRateRank_32,index_32"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/Lucifer_zzq/article/details/81236174" target="_blank" title="深度学习干货学习（1）——center loss - Lucifer_zzq的博客">
<h4 class="text-truncate oneline" style="width: 618px;">
深度学习干货学习（1）——center loss - Lucifer_zzq的博客 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/E/A/5/3_lucifer_zzq.jpg" alt="Lucifer_zzq" class="avatar-pic">
<span class="namebox">
<span class="name">Lucifer_zzq</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">07-27</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
328</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/Lucifer_zzq/article/details/81236174" target="_blank" title="深度学习干货学习（1）——center loss - Lucifer_zzq的博客">
<span class="desc oneline">在构建loss时pytorch常用的包中有最常见的MSE、cross entropy（logsoftmax+NLLLoss）、KL散度Loss、BCE、HingeLoss等等，详见：https://p...</span>
</a>
<span class="blog_title_box oneline"><a target="_blank" href="https://blog.csdn.net/Lucifer_zzq">来自： <span class="blog_title"> Lucifer_zzq的博客</span></a></span>
</p>
</div>
</div>
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/oJiMoDeYe12345/article/details/78549988,BlogCommendClickRateRank_33,index_33"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/oJiMoDeYe12345/article/details/78549988,BlogCommendClickRateRank_33,index_33"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/oJiMoDeYe12345/article/details/78549988" target="_blank" title="论文阅读学习 - Center Loss: Caffe 实现 - 长风破浪会有时，直挂云帆济沧海">
<h4 class="text-truncate oneline" style="width: 610px;">
论文阅读学习 - Center Loss: Caffe 实现 - 长风破浪会有时，直挂云帆济沧海 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/0/4/B/3_ojimodeye12345.jpg" alt="oJiMoDeYe12345" class="avatar-pic">
<span class="namebox">
<span class="name">oJiMoDeYe12345</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">11-16</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
1335</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/oJiMoDeYe12345/article/details/78549988" target="_blank" title="论文阅读学习 - Center Loss: Caffe 实现 - 长风破浪会有时，直挂云帆济沧海">
<span class="desc oneline">Center Loss - Caffe实现</span>
</a>
<span class="blog_title_box oneline"><a target="_blank" href="https://blog.csdn.net/oJiMoDeYe12345">来自： <span class="blog_title"> 长风破浪会有时，直挂云帆济沧海</span></a></span>
</p>
</div>
</div>
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/fuwenyan/article/details/72903290,BlogCommendClickRateRank_34,index_34"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/fuwenyan/article/details/72903290,BlogCommendClickRateRank_34,index_34"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/fuwenyan/article/details/72903290" target="_blank" title="Center Loss层 - VivienFu的专栏">
<h4 class="text-truncate oneline" style="width: 610px;">
Center Loss层 - VivienFu的专栏 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/D/4/5/3_fuwenyan.jpg" alt="fuwenyan" class="avatar-pic">
<span class="namebox">
<span class="name">fuwenyan</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">06-07</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
2529</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/fuwenyan/article/details/72903290" target="_blank" title="Center Loss层 - VivienFu的专栏">
<span class="desc oneline">雷锋网按：本文作者祝浩(皮搋子狐狸)，3M Cogent Beijing R&D 高级算法工程师。本硕分别毕业于哈尔滨工业大学机械专业和北京师范大学计算机专业，并于2012年加入3M。14年拿到NIC...</span>
</a>
<span class="blog_title_box oneline"><a target="_blank" href="https://blog.csdn.net/fuwenyan">来自： <span class="blog_title"> VivienFu的专栏</span></a></span>
</p>
</div>
</div>
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/wfei101/article/details/78024346,BlogCommendClickRateRank_35,index_35"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/wfei101/article/details/78024346,BlogCommendClickRateRank_35,index_35"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/wfei101/article/details/78024346" target="_blank" title="centerloss的论文解读 - BigCowPeking">
<h4 class="text-truncate oneline" style="width: 618px;">
<em>centerloss</em>的论文解读 - BigCowPeking </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/3/8/D/3_wfei101.jpg" alt="wfei101" class="avatar-pic">
<span class="namebox">
<span class="name">wfei101</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">09-18</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
612</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/wfei101/article/details/78024346" target="_blank" title="centerloss的论文解读 - BigCowPeking">
<span class="desc oneline">centerloss的论文解读</span>
</a>
<span class="blog_title_box oneline"><a target="_blank" href="https://blog.csdn.net/wfei101">来自： <span class="blog_title"> BigCowPeking</span></a></span>
</p>
</div>
</div>
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/u011918382/article/details/78832958,BlogCommendClickRateRank_36,index_36"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/u011918382/article/details/78832958,BlogCommendClickRateRank_36,index_36"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/u011918382/article/details/78832958" target="_blank" title="对Center loss的理解 - u011918382的博客">
<h4 class="text-truncate oneline" style="width: 618px;">
对Center loss的理解 - u011918382的博客 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/A/C/6/3_u011918382.jpg" alt="u011918382" class="avatar-pic">
<span class="namebox">
<span class="name">u011918382</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">12-18</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
545</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/u011918382/article/details/78832958" target="_blank" title="对Center loss的理解 - u011918382的博客">
<span class="desc oneline">在常见的目标、场景、或动作识别中，测试样本的类别包含在训练集中，这就是常说的闭集鉴定。这对softmax来说游刃有余，就像现行分类一样。然而在人脸识别领域却会出现问题，因为地球上有60亿人，模型所用训...</span>
</a>
<span class="blog_title_box oneline"><a target="_blank" href="https://blog.csdn.net/u011918382">来自： <span class="blog_title"> u011918382的博客</span></a></span>
</p>
</div>
</div>
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/App_12062011/article/details/78510783,BlogCommendClickRateRank_37,index_37"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/App_12062011/article/details/78510783,BlogCommendClickRateRank_37,index_37"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/App_12062011/article/details/78510783" target="_blank" title="人脸验证（三）--FaceNet - 工作笔记">
<h4 class="text-truncate oneline" style="width: 610px;">
<em>人脸</em><em>验证</em>（三）--FaceNet - 工作笔记 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/9/0/8/3_app_12062011.jpg" alt="App_12062011" class="avatar-pic">
<span class="namebox">
<span class="name">App_12062011</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">11-12</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
2487</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/App_12062011/article/details/78510783" target="_blank" title="人脸验证（三）--FaceNet - 工作笔记">
<span class="desc oneline">转自：blog.csdn.net/stdcoutzyx/article/details/46687471
引入
随着深度学习的出现，CV领域突破很多，甚至掀起了一股CV界的创业浪潮，当次风口浪尖之…
来自： 工作笔记
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/kwame211/article/details/79398311,BlogCommendClickRateRank_38,index_38"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/kwame211/article/details/79398311,BlogCommendClickRateRank_38,index_38"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/kwame211/article/details/79398311" target="_blank" title="centerloss系列文章笔记——人脸检测Face R-CNN - kwame211的博客">
<h4 class="text-truncate oneline" style="width: 618px;">
<em>centerloss</em>系列文章笔记——<em>人脸</em>检测Face R-CNN - kwame211的博客 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/C/0/0/3_kwame211.jpg" alt="kwame211" class="avatar-pic">
<span class="namebox">
<span class="name">kwame211</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">02-28</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
106</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/kwame211/article/details/79398311" target="_blank" title="centerloss系列文章笔记——人脸检测Face R-CNN - kwame211的博客">
<span class="desc oneline">Face R-CNN. Hao Wang Zhifeng Li∗ Xing Ji Yitong WangTencent AI Lab, ChinaFCN做人脸检测的paper，在FDDB和WIDERF...</span>
</a>
<span class="blog_title_box oneline"><a target="_blank" href="https://blog.csdn.net/kwame211">来自： <span class="blog_title"> kwame211的博客</span></a></span>
</p>
</div>
</div>
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/wfei101/article/details/79252021,BlogCommendClickRateRank_39,index_39"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/wfei101/article/details/79252021,BlogCommendClickRateRank_39,index_39"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/wfei101/article/details/79252021" target="_blank" title="目标检测：损失函数之SmoothL1Loss - BigCowPeking">
<h4 class="text-truncate oneline" style="width: 610px;">
目标检测：损失函数之SmoothL1Loss - BigCowPeking </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/3/8/D/3_wfei101.jpg" alt="wfei101" class="avatar-pic">
<span class="namebox">
<span class="name">wfei101</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">02-04</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
6073</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/wfei101/article/details/79252021" target="_blank" title="目标检测：损失函数之SmoothL1Loss - BigCowPeking">
<span class="desc oneline">fasterRCNN中的RPN的回归框的loss计算方法
再次深入的解读SmoothL1Loss损失函数：
fasterRCNN中SmoothL1Loss的详解
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/YLH9604/article/details/74295095,BlogCommendClickRateRank_40,index_40"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/YLH9604/article/details/74295095,BlogCommendClickRateRank_40,index_40"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/YLH9604/article/details/74295095" target="_blank" title="损失函数 center_loss triplet loss - YLH9604的专栏">
<h4 class="text-truncate oneline" style="width: 610px;">
损失函数 center_loss triplet loss - YLH9604的专栏 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/5/D/0/3_ylh9604.jpg" alt="YLH9604" class="avatar-pic">
<span class="namebox">
<span class="name">YLH9604</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">07-04</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
2420</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/YLH9604/article/details/74295095" target="_blank" title="损失函数 center_loss triplet loss - YLH9604的专栏">
<span class="desc oneline">A Discriminative Feature Learning Approach for Deep Face Recognition 原理及在caffe实验复现
图像特征提取系列之ＰＣ…
来自： YLH9604的专栏
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/u011808673/article/details/80345772,BlogCommendClickRateRank_41,index_41"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/u011808673/article/details/80345772,BlogCommendClickRateRank_41,index_41"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/u011808673/article/details/80345772" target="_blank" title="近来人脸识别的 loss - Nine days">
<h4 class="text-truncate oneline" style="width: 618px;">
近来<em>人脸</em>识别的 loss - Nine days </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/6/7/3/3_u011808673.jpg" alt="u011808673" class="avatar-pic">
<span class="namebox">
<span class="name">u011808673</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">05-17</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
589</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/u011808673/article/details/80345772" target="_blank" title="近来人脸识别的 loss - Nine days">
<span class="desc oneline">最近（2017/2018）人脸识别的相关论文人脸识别的Loss函数改进的论文比较多, 如:[2017] L2-constrained Softmax Loss for Discriminative F...</span>
</a>
<span class="blog_title_box oneline"><a target="_blank" href="https://blog.csdn.net/u011808673">来自： <span class="blog_title"> Nine days</span></a></span>
</p>
</div>
</div>
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/marleylee/article/details/81212669,BlogCommendClickRateRank_42,index_42"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/marleylee/article/details/81212669,BlogCommendClickRateRank_42,index_42"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/marleylee/article/details/81212669" target="_blank" title="人脸识别之人脸验证（五）--NormFace - marleylee的博客">
<h4 class="text-truncate oneline" style="width: 626px;">
<em>人脸</em>识别之<em>人脸</em><em>验证</em>（五）--NormFace - marleylee的博客 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/B/B/9/3_marleylee.jpg" alt="marleylee" class="avatar-pic">
<span class="namebox">
<span class="name">marleylee</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">07-26</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
88</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/marleylee/article/details/81212669" target="_blank" title="人脸识别之人脸验证（五）--NormFace - marleylee的博客">
<span class="desc oneline">转自：http://blog.csdn.net/sinat_27614125/article/details/71104183?locationNum=16&amp;amp;fps=1 略删改。
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/tinyzhao/article/details/53015724,BlogCommendClickRateRank_43,index_43"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/tinyzhao/article/details/53015724,BlogCommendClickRateRank_43,index_43"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/tinyzhao/article/details/53015724" target="_blank" title="人脸验证：Joint Bayesian - flyingzhao">
<h4 class="text-truncate oneline" style="width: 610px;">
<em>人脸</em><em>验证</em>：Joint Bayesian - flyingzhao </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/7/3/0/3_tinyzhao.jpg" alt="tinyzhao" class="avatar-pic">
<span class="namebox">
<span class="name">tinyzhao</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">11-03</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
6982</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/tinyzhao/article/details/53015724" target="_blank" title="人脸验证：Joint Bayesian - flyingzhao">
<span class="desc oneline">《Bayesian Face Revisited: A Joint Formulation》论文解读这篇文章发表于ECCV2012，来自MSRA的孙剑组。论文原文见：
http://home.us…
来自： flyingzhao
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/wuzuyu365/article/details/51790002,BlogCommendClickRateRank_44,index_44"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/wuzuyu365/article/details/51790002,BlogCommendClickRateRank_44,index_44"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/wuzuyu365/article/details/51790002" target="_blank" title="一个不错的深度学习人脸验证 - walle的博客">
<h4 class="text-truncate oneline" style="width: 618px;">
一个不错的深度学习<em>人脸</em><em>验证</em> - walle的博客 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/8/E/7/3_wuzuyu365.jpg" alt="wuzuyu365" class="avatar-pic">
<span class="namebox">
<span class="name">wuzuyu365</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">06-30</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
631</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/wuzuyu365/article/details/51790002" target="_blank" title="一个不错的深度学习人脸验证 - walle的博客">
<span class="desc oneline">https://github.com/happynear</span>
</a>
<span class="blog_title_box oneline"><a target="_blank" href="https://blog.csdn.net/wuzuyu365">来自： <span class="blog_title"> walle的博客</span></a></span>
</p>
</div>
</div>
<div class="recommend-item-box recommend-box-ident recommend-download-box clearfix" data-track-view="{"mod":"popu_614","con":",https://download.csdn.net/download/questioncat/9657585,BlogCommendClickRateRank_45,index_45"}" data-track-click="{"mod":"popu_614","con":",https://download.csdn.net/download/questioncat/9657585,BlogCommendClickRateRank_45,index_45"}" data-flg="true">
<a href="https://download.csdn.net/download/questioncat/9657585" target="_blank">
<div class="content">
<div class="">
<span class="type">下载</span>
<h4 class="text-truncate oneline clearfix">
<em>人脸</em><em>验证</em>Face-Verification </h4>
<span class="data float-right">10-19</span>
</div>
<div class="desc oneline">
人脸验证C代码 </div>
</div>
</a>
</div>
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/dongfang1984/article/details/53337195,BlogCommendClickRateRank_46,index_46"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/dongfang1984/article/details/53337195,BlogCommendClickRateRank_46,index_46"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/dongfang1984/article/details/53337195" target="_blank" title="A Discriminative Feature Learning Approach for Deep Face Recognition 原理及在caffe实验复现 - dongfang1984的博客">
<h4 class="text-truncate oneline" style="width: 608px;">
A Discriminative Feature Learning Approach for Deep Face Recognition 原理及在caffe实验复现 - dongfang1984的博客 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/4/D/3/3_dongfang1984.jpg" alt="dongfang1984" class="avatar-pic">
<span class="namebox">
<span class="name">dongfang1984</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">11-25</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
1.2万</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/dongfang1984/article/details/53337195" target="_blank" title="A Discriminative Feature Learning Approach for Deep Face Recognition 原理及在caffe实验复现 - dongfang1984的博客">
<span class="desc oneline">本文主要讲centerloss 的原理，及其创新点。然后用caffe 跑自己的数据（CASIA-WebFace | MsCelebV1-Faces-Aligned）
Reference pap…
来自： dongfang1984的博客
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/wuzuyu365/article/details/73530622,BlogCommendClickRateRank_47,index_47"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/wuzuyu365/article/details/73530622,BlogCommendClickRateRank_47,index_47"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/wuzuyu365/article/details/73530622" target="_blank" title="采用center loss的caffe 人脸识别 - walle的博客">
<h4 class="text-truncate oneline" style="width: 610px;">
采用center loss的caffe <em>人脸</em>识别 - walle的博客 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/8/E/7/3_wuzuyu365.jpg" alt="wuzuyu365" class="avatar-pic">
<span class="namebox">
<span class="name">wuzuyu365</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">06-21</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
2666</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/wuzuyu365/article/details/73530622" target="_blank" title="采用center loss的caffe 人脸识别 - walle的博客">
<span class="desc oneline">https://sanwen8.cn/p/446HEbm.html</span>
</a>
<span class="blog_title_box oneline"><a target="_blank" href="https://blog.csdn.net/wuzuyu365">来自： <span class="blog_title"> walle的博客</span></a></span>
</p>
</div>
</div>
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/samylee/article/details/76640240,BlogCommendClickRateRank_48,index_48"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/samylee/article/details/76640240,BlogCommendClickRateRank_48,index_48"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/samylee/article/details/76640240" target="_blank" title="原始caffe添加CenterLoss_layer - samylee的博客">
<h4 class="text-truncate oneline" style="width: 610px;">
原始caffe添加<em>CenterLoss</em>_layer - samylee的博客 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/2/E/E/3_samylee.jpg" alt="samylee" class="avatar-pic">
<span class="namebox">
<span class="name">samylee</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">08-03</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
2140</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/samylee/article/details/76640240" target="_blank" title="原始caffe添加CenterLoss_layer - samylee的博客">
<span class="desc oneline">CenterLoss_layer可以在原分类的基础上（某种程度上）可提升几个点左右（博主测试提升6个点），还不占前向时间，好东西啊！
原理这里不介绍了，网上应有尽有！这里只是介绍如何在caffe中如何…
来自： samylee的博客
<div class="recommend-item-box recommend-box-ident type_blog clearfix" data-track-view="{"mod":"popu_614","con":",https://blog.csdn.net/zbzb1000/article/details/81001546,BlogCommendClickRateRank_49,index_49"}" data-track-click="{"mod":"popu_614","con":",https://blog.csdn.net/zbzb1000/article/details/81001546,BlogCommendClickRateRank_49,index_49"}" data-flg="true">
<div class="content" style="width: 702px;">
<a href="https://blog.csdn.net/zbzb1000/article/details/81001546" target="_blank" title="CenterLoss的caffe实现与深入理解 - AlphaHolo的博客">
<h4 class="text-truncate oneline" style="width: 618px;">
<em>CenterLoss</em>的caffe实现与深入理解 - AlphaHolo的博客 </h4>
<div class="info-box d-flex align-content-center">
<!-- <p class="avatar">
<img src="https://avatar.csdn.net/9/D/0/3_zbzb1000.jpg" alt="zbzb1000" class="avatar-pic">
<span class="namebox">
<span class="name">zbzb1000</span>
<span class="triangle"></span>
</span>
</p> -->
<p class="date-and-readNum">
<span class="date hover-show">07-11</span>
<span class="read-num hover-hide">
<svg class="icon csdnc-yuedushu" aria-hidden="true">
<use xlink:href="#csdnc-m-passwords-visible"></use>
</svg>
407</span>
</p>
</div>
</a>
<p class="content" style="width: 702px;">
<a href="https://blog.csdn.net/zbzb1000/article/details/81001546" target="_blank" title="CenterLoss的caffe实现与深入理解 - AlphaHolo的博客">
<span class="desc oneline">CenterLoss的caffe实现与深入理解
目的：
打算将centerloss与SSD的multibox_loss融合，或者与faster-rcnn的softmax回归融合。
项目代码将…
来自： AlphaHolo的博客
<div class="recommend-loading-box">
<img src="https://csdnimg.cn/release/phoenix/images/feedLoading.gif">
</div>
<div class="recommend-end-box">
<p class="text-center">没有更多推荐了，<a href="https://blog.csdn.net/" class="c-blue c-blue-hover c-blue-focus">返回首页</a></p>
</div>
</div>
</main>
有点纠结，实质上本文只是在深度学习框架下人脸识别损失函数的改进。但我还是把它归类到端到端了。出于个人原因。