Efficient Dense 3D Rigid-Body Motion Segmentation in RGB-D Video thumbnail
Pause
Mute
Subtitles
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

Efficient Dense 3D Rigid-Body Motion Segmentation in RGB-D Video

Published on Apr 03, 20142508 Views

Motion is a fundamental segmentation cue in video. Many current approaches segment 3D motion in monocular or stereo image sequences, mostly relying on sparse interest points or being dense but compu

Related categories

Chapter list

Efficient Dense 3D Rigid-Body Motion Segmentation in RGB-D Video00:00
Objective - 100:00
Objective - 200:18
Objective - 300:31
Approach - 100:51
Approach - 201:00
Approach - 301:04
Unary potentials and pair-wise potentials01:10
Discount labelling-pairs01:34
Back to the EM-Objective02:00
Factored Observations02:13
Variational Approximation - 102:26
Variational Approximation - 202:48
Minimizing KL-divergence03:20
Initialize mean-field iterations - 103:58
Initialize mean-field iterations - 204:08
Trade model complexity via label costs04:29
Efficient EM-Algorithm: E-Step - 105:38
Efficient EM-Algorithm: M-Step06:39
Efficient EM-Algorithm: E-Step - 206:55
Sequential EM07:12
Efficient Image Representation07:36
Experiments: Setup08:32
Experiments: Results - 109:27
Experiments: Results - 211:20
Summary11:41