好文作者面授招-20140916-Ensemble-based tracking: Aggregating crowdsourced structured time series data

  • 主题:Ensemble-based tracking: Aggregating crowdsourced structured time series data
  • 主讲人王乃岩
  • 主持人孟德宇
  • 活动时间:2014年9月16日星期2,北京时间20:00开始。

1. 相关文献阅读

  • Naiyan Wang, Dit-Yan Yeung. Ensemble-based tracking: Aggregating crowdsourced structured time series data. in Proceedings of the Thirty-First International Conference on Machine Learning (ICML) , Beijing, China, 21-26 June 2014.[pdf]

2. 讲座资料

3. 代码数据下载

4. 精彩问答集锦

Locations of visitors to this page

17 thoughts on “好文作者面授招-20140916-Ensemble-based tracking: Aggregating crowdsourced structured time series data

  • September 16, 2014 at 9:37 pm
    Permalink

    您的算法ensemble了多少tracker,这样的话会不会影响tracking的速度。什么样的配置?

    Reply
  • September 16, 2014 at 9:37 pm
    Permalink

    我有个问题,如果一个tracker是正确的,其他tracker都是错误的,那combine之后岂不是要把这个结果扔掉吗?

    Reply
  • September 16, 2014 at 9:37 pm
    Permalink

    另外这个时间是不是应该算上其他tracker的时间呢?

    Reply
  • September 16, 2014 at 9:38 pm
    Permalink

    视频质量对tracking影响多大?

    Reply
  • September 16, 2014 at 9:38 pm
    Permalink

    5FPS这个速度 ?C++呢?

    Reply
  • September 16, 2014 at 9:39 pm
    Permalink

    时序的model ensemble可以推广到action之类的问题的吧?

    Reply
  • September 16, 2014 at 9:40 pm
    Permalink

    我理解是每个tracker适应的视频bias不同,然后selector实际上可以看成学习了当时刻的bias?

    Reply
  • September 16, 2014 at 9:40 pm
    Permalink

    如果我只有少数几个tracker的算法(CSK),我能不能够做出几个不同bias的response让他们average来得到更好的结果

    Reply
  • September 16, 2014 at 9:40 pm
    Permalink

    有没有单独分析过遮挡、光线、视角等条件变化下,哪些TRACKER更好吗?

    Reply
  • September 16, 2014 at 9:41 pm
    Permalink

    你的论文和屠卓文的论文的本质区别在哪 Quannan Li, Xinggang Wang, Wei Wang, Yuan Jiang, Zhi-Hua Zhou, and Zhuowen Tu, “Disagreement-Based Multi-System Tracking”, Workshop of Detection and Tracking in Challenging Environments, 2012.

    Reply
  • September 16, 2014 at 9:41 pm
    Permalink

    对于丢失问题,处理结果怎样?

    Reply
  • September 16, 2014 at 9:41 pm
    Permalink

    Tracker的阈值是变化的吗?还是一直是固定的

    Reply
  • September 16, 2014 at 9:41 pm
    Permalink

    跟踪的结果是综合的所有trackers的结果,如何保证这个结果更接近ground truth?

    Reply
  • September 16, 2014 at 9:41 pm
    Permalink

    Tracking 中的漂移问题,本文的方法是怎么克服的?

    Reply
  • September 16, 2014 at 9:41 pm
    Permalink

    文章提到了WHOLE TEMPLATE 方法和LOCAL PATCH方法,这个方法有可能直接将多个TRACKER初始成不同PATCH做成PATCH-BASED METHOD吗?

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *