0.25
0.5
0.75
1.25
1.5
1.75
2
Normalized Cut meets MRF
Published on Oct 24, 20162632 Views
We propose a new segmentation model combining common regularization energies, e.g. Markov Random Field (MRF) potentials, and standard pairwise clustering criteria like Normalized Cut (NC), average ass
Related categories
Chapter list
Normalized Cut Meets MRF00:00
Normalized Cut (NC) - 100:11
Normalized Cut (NC) - 200:31
Markov Random Field (MRF) - 100:39
Markov Random Field (MRF) - 200:50
Markov Random Field (MRF) - 300:59
Markov Random Field (MRF) - 401:08
Our proposal: Normalized Cut + MRF01:15
Why MRF for Normalized Cut? - 101:32
Why MRF for Normalized Cut? - 202:12
Why MRF for Normalized Cut? - 302:40
Why MRF for Normalized Cut? - 403:32
Our proposal: Normalized Cut + MRF04:23
Bound optimization, in general05:05
Bound for our joint energy05:30
Kernel bound for NC05:53
Low-rank Approximation06:34
NC - 107:39
NC08:23
NC - 309:08
Our Kernel Cut and Spectral Cut09:30
Experiments: MRF helps Normalized Cut - 109:54
Experiments: MRF helps Normalized Cut - 210:25
Experiments: Normalized Cut helps MRF10:37
More Experiments11:12
Conclusion11:18