人脸对齐算法常用评价标准

<span class="tags-box artic-tag-box">
<span class="label">标签：</span>
<a data-track-click="{&quot;mod&quot;:&quot;popu_626&quot;,&quot;con&quot;:&quot;Face Alignment&quot;}" class="tag-link" href="http://so.csdn.net/so/search/s.do?q=Face Alignment&amp;t=blog" target="_blank">Face Alignment																</a><a data-track-click="{&quot;mod&quot;:&quot;popu_626&quot;,&quot;con&quot;:&quot;评估标准&quot;}" class="tag-link" href="http://so.csdn.net/so/search/s.do?q=评估标准&amp;t=blog" target="_blank">评估标准																</a>
<span class="article_info_click">更多</span></span>
<div class="tags-box space">
<span class="label">个人分类：</span>
<a class="tag-link" href="https://blog.csdn.net/u011681952/article/category/7963613" target="_blank">Face Alignment																</a>
</div>
</div>
<div class="operating">
</div>
</div>
</div>
</div>
<article class="baidu_pl">
<div id="article_content" class="article_content clearfix csdn-tracking-statistics" data-pid="blog" data-mod="popu_307" data-dsm="post">
<div class="article-copyright">
版权声明：未经博主允许，不得转载！					https://blog.csdn.net/u011681952/article/details/81985559				</div>
<div class="markdown_views prism-atom-one-dark">
<!-- flowchart 箭头图标 勿删 -->
<svg xmlns="http://www.w3.org/2000/svg" style="display: none;"><path stroke-linecap="round" d="M5,0 0,2.5 5,5z" id="raphael-marker-block" style="-webkit-tap-highlight-color: rgba(0, 0, 0, 0);"></path></svg>
<p><strong>人脸对齐常用评价标准</strong></p>

1 the inter-ocular distance normalized error

       
       其中xixi也可以用两外眼角间距离、人脸shape外接矩形对角线长度代替

2 the mean normalized error (MNE)

       
       符号意义与上面相同，仅仅多了一项平均项

3 the area-under-the-curve (AUCaAUCa)

       AUCaAUCa值不像平均误差那样受单个点误差较大而发生较大变化，非常敏感
       详见：https://www.geek-share.com/detail/2731412153.html

4 the Cumulative Errors Distribution (CED) curve

       CED与AUC有些类似，但又不同；如下图（来自Look at Boundary论文）
       
       横坐标表示归一化点与点间误差ee时，LAB的失败率为2.17%；另外图中Error是指平均误差（MNE），见方法2