当前位置:首页 > 学术论坛

【物流讲坛】Visual Objects Tracking: Challenges and Opportunities

发布时间:2017-06-21

   

  主讲人:胡玉衡 

  时间:2017628日下午2:30 

  地点:物流学院学术报告厅218 

  主题:Visual Objects Tracking: Challenges and Opportunities 

    

  YuHen Hu,胡玉衡,1976 年毕业于台湾国立大学,分别在1979 年和1982 年于南加利福尼亚州大学获得硕士及博士学位。现为美国威斯康星大学麦迪逊分校(University of WisconsinMadison)电机与计算机工程系副主任、教授。胡玉衡教授是IEEE 院士,IEEE Transaction of AcousticSpeechand Signal Processing 等多个杂志编委及IEEE 多个国际会议主席,现为美国威斯康星大学麦迪逊分校(University of Wisconsin-Madison)电机与计算机工程系副主任、教授, 目前研究领域为:信号处理算法的设计与实现,VLSI 的计算机辅助设计和物理设计,模式分类和机器学习,人工神经网络及其工程应用,图像与信号处理。 

  胡玉衡教授是电子电机工程学会的会士(Fellow of Institute of Electrical and Electronics Engineers, IEEE),曾任IEEE 信号处理协会(Signal Processing Society)秘书长,并代表该协会担任IEEE 类神经网络理事会(Neural Network Council)董事。在传感网研究领域久负盛名,主要研究涉及信号处理算法的设计和实现、大规模集成电路计算机辅助设计、模式识别和机器学习算法、图像信号处理等领域,发表论文200 余篇,并编写多部专著。 

  ABSTRACT 

  Visual objects tracking is a key enabling technology for numerous emerging applications ranging from security surveillance, intelligent transportation, smart city, connected health care, to name a few. In this presentation, an overview of the visual object tracking problem will be given. Major technical challenges, including non-rigid visual object detection and localization, structured and non-planar motion modeling, as well as multiple target tracking will be surveyed and recent progress will be discussed. Recent progress of novel VOT techniques, such as deep learning will be surveyed. Emerging video capturing instrumentations, such as camera array, and associated VoT techniques will be reviewed.