文献知网节
  • 记笔记

Real-Time Multi-Camera Multi-Person Action Recognition using Pose Estimation

Jonathan Then Sien PhangKing Hann Lim

Curtin University Malaysia

摘要:Action recognition possesses challenging issues in real-time multi-camera scenario when dealing with multi-person such as occlusion, pose variance and action interaction. In this paper, a real-time pipeline is proposed to address multi-person action recognition in multi-camera setup using joint key-points sequences from detected person. Joints trajectory is the important time-series information to identify actions. 14 key-points from human joints are scaled with relative to the Euclidean distance of neck-to-pelvis to obtain standard size of person, which is invariant to camera distance. Subsequently, 3 D histogram correlation is applied to match multi-person identity. An indexed person with a series of action attribute are collected and fed into Long ShortTerm Memory(LSTM) recurrent neural network. The proposed pipeline uses spatial-temporal feature of person’s joint key-points trajectory for action recognition. Minimal single pass forward time through the LSTM network enables real-time multi-person action recognition in a video sequence. The proposed pipeline achieved up to 13 frames per second with 92% recognition rate with two camera setups.
会议名称:

The 3rd International Conference on Machine Learning and Soft Computing (ICMLSC 2019)

会议时间:

2019-01-25

会议地点:

Da Lat,Vietnam

  • 专辑:

    电子技术及信息科学

  • 专题:

    计算机软件及计算机应用

  • DOI:

    10.26914/c.cnkihy.2019.042391

  • 分类号:

    TP391.41

  • 手机阅读
    即刻使用手机阅读
    第一步

    扫描二维码下载

    "移动知网-全球学术快报"客户端

    第二步

    打开“全球学术快报”

    点击首页左上角的扫描图标

    第三步

    扫描二维码

    手机同步阅读本篇文献

  • CAJ下载
  • PDF下载

下载手机APP用APP扫此码同步阅读该篇文章

下载:6 页码:182-187 页数:6 大小:726k

相关推荐
  • 相似文献
  • 读者推荐
  • 相关基金文献
  • 关联作者
  • 相关视频