查看“动捕小组/计算机视觉”的源代码

= 主干领域 =
<html>
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<meta http-equiv="X-UA-Compatible" content="ie=edge">
<title>Markmap</title>
<style>
* {
  margin: 0;
  padding: 0;
}
#mindmap {
  display: block;
  width: 47vw;
  height: 25vh;
}
</style>
<link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/markmap-toolbar@0.15.3/dist/style.css">
</head>
<body>
<svg id="mindmap"></svg>
<script src="https://cdn.jsdelivr.net/npm/d3@7.8.5/dist/d3.min.js"></script><script src="https://cdn.jsdelivr.net/npm/markmap-view@0.15.3/dist/browser/index.js"></script><script src="https://cdn.jsdelivr.net/npm/markmap-toolbar@0.15.3/dist/index.js"></script><script>(r => {
                setTimeout(r);
              })(() => {
  const {
    markmap,
    mm
  } = window;
  const {
    el
  } = markmap.Toolbar.create(mm);
  el.setAttribute('style', 'position:absolute;bottom:20px;right:20px');
})</script><script>((getMarkmap, getOptions, root, jsonOptions) => {
          const markmap = getMarkmap();
          window.mm = markmap.Markmap.create('svg#mindmap', (getOptions || markmap.deriveOptions)(jsonOptions), root);
        })(() => window.markmap,null,{"type":"heading","depth":0,"payload":{"lines":[0,1]},"content":"计算机视觉","children":[{"type":"heading","depth":1,"payload":{"lines":[1,2]},"content":"人体姿态估计","children":[{"type":"heading","depth":2,"payload":{"lines":[2,3]},"content":"二维","children":[]},{"type":"heading","depth":2,"payload":{"lines":[3,4]},"content":"三维","children":[]}]},{"type":"heading","depth":1,"payload":{"lines":[4,5]},"content":"基于单目视频","children":[{"type":"heading","depth":2,"payload":{"lines":[5,6]},"content":"帕金森患者UPDRS评分预测","children":[]},{"type":"heading","depth":2,"payload":{"lines":[6,7]},"content":"人体运动分析","children":[]}]}]},{})</script>
</body>
</html>
= 人体姿态估计 =
== 二维 ==
{{文章展开|name=OpenPose: Realtime Multi-Person 2D Pose Estimation using Part Affinity Fields|prefix=动捕小组/计算机视觉|toggled=二维人体姿态估计方法领域的经典文章，提出了多人实时二维姿态估计方法，在计算资源可观的情况下兼具实时性和较高的准确度，在该领域内的相关研究应用中被广泛引用。|Citation_=11783}}
{{文章展开|name=BlazePose: On-device Real-time Body Pose Tracking|prefix=动捕小组/计算机视觉|toggled=轻量级二维人体姿态估计经典方法，具有算力资源要求低、准确度高等特点，可部署在笔记本电脑、移动手机等边缘设备上，当下被广泛应用于使用轻量级二维人体姿态估计方法实现各类运动分析的应用中。|Citation_=467}}
== 三维 ==
{{文章展开|name=3D Human pose estimation in video with temporal convolutions and semi-supervised training|prefix=动捕小组/计算机视觉|toggled=基于单目视频实现三维人体姿态估计的经典方法，文中在时序上对多帧图像的二维人体关键点使用空洞时域卷积，较为准确地估计当前帧的人体姿态；并应用半监督学习的方法，有效利用了不含标签的视频数据训练模型。|Citation_=1005}}
= 基于单目视频 =
== 帕金森患者UPDRS评分预测 ==
{{文章展开|name=Quantifying Parkinson’s Disease Motor Severity Under Uncertainty Using MDS-UPDRS Videos|prefix=动捕小组/计算机视觉|toggled=本文提出评分混淆矩阵解决不同评分者对同一帕金森病患者的MDS-UPDRS视频有不同评分导致的标签噪声问题，并设计了神经网络模型对输入的MDS-UPDRS视频预测分类其MDS-UPDRS评分，是计算机视觉技术在跟踪评估帕金森患者病重程度方向的实际应用示例。|Citation_=44}}
== 人体运动分析 ==
{{文章展开|name=Deep neural networks enable quantitative movement analysis using single-camera videos|prefix=动捕小组/计算机视觉|toggled=本文提出的方法使得基于居家或于医院使用普通相机拍摄的单目视频进行人体的量化运动分析成为可能，而不再依赖于深度相机、光学动捕设备等昂贵硬件采集数据，可参考本文中提出的方法流程设计便捷的分析方法开展大规模神经肌肉类疾病患者的量化运动分析。|Citation_=175}}
{{主领域分类}}