published on 2014-01-15
2346 views
  • 00:51 1.
    Tracking Human Under Occlusion Based on Adaptive Multiple Kernels With Projected GradientsIEEE TRANSACTIONS ON MULTIMEDIA, VOL.15, NO.7 NOVEMBER 2013
  • 00:33 2.
    Outline
  • 00:11 3.
    Introduction
  • 00:45 4.
    Introduction
  • 00:03 5.
    Point Tracking
  • 00:01 6.
    Introduction
  • 00:24 7.
    Point Tracking
  • 00:34 8.
    Kernel Tracking
  • 00:02 9.
    Point Tracking
  • 00:13 10.
    Silhouette Tracking
  • 01:10 11.
    Main Contributions
  • 00:05 12.
    Proposed Method
  • 00:51 13.
    Multiple-Kernel Tracking
  • 00:01 14.
    Single Kernel Tracking
  • 00:01 15.
    Multiple-Kernel Cost Formulation
  • 05:14 16.
    Multiple-Kernel Cost Formulation
  • 00:01 17.
    Single Kernel Tracking
  • 00:01 18.
    Multiple-Kernel Tracking
  • 00:42 19.
    Multiple-Kernel Tracking
  • 00:05 20.
    Single Kernel Tracking
  • 00:35 21.
    Single Kernel Tracking
  • 00:47 22.
    Single Kernel Tracking
  • 00:01 23.
    Single Kernel Tracking
  • 01:55 24.
    Single Kernel Tracking
  • 00:14 25.
    Single Kernel Tracking
  • 00:17 26.
    Single Kernel Tracking
  • 00:04 27.
    Single Kernel Tracking
  • 00:02 28.
    Single Kernel Tracking
  • 01:41 29.
    Single Kernel Tracking
  • 00:39 30.
    Multiple-Kernel Cost Formulation
  • 00:01 31.
    Multiple-Kernel Cost Formulation
  • 00:10 32.
    Multiple-Kernel Cost Formulation
  • 01:13 33.
    Multiple-Kernel Cost Formulation
  • 00:01 34.
    Multiple-Kernel Cost Formulation
  • 00:27 35.
    Multiple-Kernel Cost Formulation
  • 00:39 36.
    Multiple-Kernel Cost Formulation
  • 00:08 37.
    Multiple-Kernel Cost Formulation
  • 00:08 38.
    Multiple-Kernel Cost Formulation
  • 00:25 39.
    Multiple-Kernel Cost Formulation
  • 00:01 40.
    Multiple-Kernel Cost Formulation
  • 00:40 41.
    Multiple-Kernel Cost Formulation
  • 00:06 42.
    Multiple-Kernel Cost Formulation
  • 00:41 43.
    Multiple-Kernel Cost Formulation
  • 00:21 44.
    Multiple-Kernel Cost Formulation
  • 00:02 45.
    Multiple-Kernel Cost Formulation
  • 00:06 46.
    Multiple-Kernel Cost Formulation
  • 00:48 47.
    Multiple-Kernel Cost Formulation
  • 00:56 48.
    Multiple-Kernel Cost Formulation
  • 02:23 49.
    Multiple-Kernel Cost Formulation
  • 00:02 50.
    Multiple-Kernel Cost Formulation
  • 00:07 51.
    Multiple-Kernel Cost Formulation
  • 03:17 52.
    Multiple-Kernel Cost Formulation
  • 00:04 53.
    Multiple-Kernel Cost Formulation
  • 00:01 54.
    Multiple-Kernel Cost Formulation
  • 00:01 55.
    Multiple-Kernel Cost Formulation
  • 00:19 56.
    Multiple-Kernel Cost Formulation
  • 00:02 57.
    Multiple-Kernel Cost Formulation
  • 00:22 58.
    Multiple-Kernel Cost Formulation
  • 00:38 59.
    Adaptive Cost Function
  • 00:30 60.
    Scale Change Issue
  • 00:41 61.
    Scale Change Issue
  • 01:34 62.
    Scale Change Issue
  • 00:04 63.
    Experimental Result
  • 00:48 64.
    Implementation Setup
  • 00:54 65.
    Implementation Setup
  • 00:04 66.
    Implementation Setup
  • 00:02 67.
    Implementation Setup
  • 00:01 68.
    Implementation Setup
  • 00:10 69.
    Implementation Setup
  • 01:21 70.
    Implementation Setup
  • 00:03 71.
    Implementation Setup
  • 01:27 72.
    Implementation Setup
  • 00:21 73.
    Experiments on Multiple-Kernel Tracking
  • 00:57 74.
    Experiments on Multiple-Kernel Tracking
  • 00:07 75.
    Experiments on Multiple-Kernel Tracking
  • 00:30 76.
    Experiments on Multiple-Kernel Tracking
  • 00:02 77.
    Experiments on Multiple-Kernel Tracking
  • 00:04 78.
    Experiments on Multiple-Kernel Tracking
  • 00:01 79.
    Experiments on Multiple-Kernel Tracking
  • 00:08 80.
    Experiments on Multiple-Kernel Tracking
  • 00:07 81.
    Experiments on Multiple-Kernel Tracking
  • 00:01 82.
    Experiments on Multiple-Kernel Tracking
  • 00:31 83.
    Experiments on Multiple-Kernel Tracking
  • 00:08 84.
    Experiments on Scale Change
  • 02:49 85.
    Experiments on the Automatic Tracking System
  • 00:19 86.
    Experiments on the Automatic Tracking System
  • 00:06 87.
    Discussion
  • 00:54 88.
    Kernel Design
  • 00:11 89.
    Kernel Design
  • 00:05 90.
    Kernel Design
  • 00:18 91.
    Kernel Design
  • 00:04 92.
    Conclusion
  • 01:08 93.
    Conclusion
  • Outline
  • Note
  • Discuss
  • Fullscreen
Tracking Human Under Occlusion Based on Adaptive Multiple Kernels With Projected Gradients
49:27, published on 2014-01-15 by VCLab NTHU