The Prague Texture Segmentation Datagenerator and Benchmark - Results
optimized
for

Results are sorted by criterion 'NMI' (normalized mutual information) in descending order. The sorting criterium or the order can be changed by clicking on the appropriate criterium label. Subset of compared results can be filtered by several filters below. The segmentation details (single mosaics and their corresponding criteria values) are visible by clicking on the order number. Below the criterium labels are mean and standard deviation of the values of displayed results. They are used to compute z-scores which are displayed in brackets. The ranks are displayed in parentheses.

Select  result's set    benchmark     hidden results Show Hide
Filter by algorithm features
? 0 1   f1 = classification (supervised segmentation)
? 0 1   f2 = hiearchy result (manual selection)
? 0 1   f3 = known number of regions
? 0 1   f4 = [reserved]
Filter by degradationtype: ? no Gaussian Poisson Salt&pepper Blur (Gaussian) Median
any 0.5 0.324 0.21 0.136 0.088 0.057 0.037 0.024 0.015 0.01
Filter by algorithm
Algorithm: Version:

   
Benchmark - Colour f1 f2 f3 f4 AVG
RANK
NORM
CS
61.46
±2.82
OS
11.98
±1.61
US
7.80
±2.85
ME
15.69
±1.98
NE
16.05
±2.11
O
11.04
±0.18
C
10.79
±0.98
CA
74.69
±0.95
CO
81.91
±0.72
CC
86.75
±1.04
I
18.09
±0.72
II
2.76
±0.25
EA
82.28
±0.94
MS
72.86
±1.08
RM
5.29
±0.29
CI
83.27
±0.89
GCE
16.57
±0.77
LCE
9.51
±0.95
BCE
27.62
±0.78
GBCE
20.56
±0.79
BGM
81.91
±0.72
SC
75.71
±0.79
SSC
76.15
±0.78
VD
13.04
±0.57
L
81.37
±0.52
AVI
5.72
±0.25
NVI
19.20
±0.84
criterion 'NMI' (normalized mutual information) in descending orderNMI
78.92
±0.86
M
8.96
±0.34
ARI
73.44
±0.94
JC
67.20
±1.12
DC
79.08
±0.76
FMI
79.35
±0.75
WI
80.51
±1.11
WII
78.75
±1.15
NBDE
5.91
±0.01
 1.  felipecalderero's GSRM sup. KL area-weighted  [normal] S(0.324) 0 0 1 0  83.27 
 (1.22) 
 [0.935] 
62.94
(2)
[0.527]
14.25
(3)
[1.413]
6.95
(2)
[-0.297]
13.15
(1)
[-1.281]
13.35
(1)
[-1.283]
10.80
(1)
[-1.280]
11.24
(2)
[0.455]
75.88
(1)
[1.248]
82.77
(1)
[1.194]
87.32
(2)
[0.540]
17.23
(1)
[-1.194]
2.49
(1)
[-1.079]
83.16
(1)
[0.937]
74.15
(1)
[1.194]
4.93
(1)
[-1.228]
84.07
(1)
[0.897]
15.86
(1)
[-0.911]
8.82
(1)
[-0.730]
26.51
(1)
[-1.409]
19.47
(1)
[-1.387]
82.77
(1)
[1.194]
76.83
(1)
[1.414]
77.25
(1)
[1.408]
12.24
(1)
[-1.411]
81.82
(1)
[0.875]
5.52
(1)
[-0.814]
18.53
(1)
[-0.798]
79.70
(1)
[0.916]
8.48
(1)
[-1.409]
74.71
(1)
[1.345]
68.62
(1)
[1.262]
80.05
(1)
[1.280]
80.28
(1)
[1.252]
81.99
(1)
[1.329]
79.07
(2)
[0.273]
5.91
(2)
[0.058]
 2.  felipecalderero's GSRM sup. KL area-weighted  [normal] S(0.21) 0 0 1 0  81.86 
 (2.47) 
 [-0.506] 
57.50
(3)
[-1.400]
10.94
(2)
[-0.652]
11.63
(3)
[1.346]
17.99
(3)
[1.160]
18.48
(3)
[1.156]
11.06
(2)
[0.119]
9.43
(1)
[-1.387]
73.55
(3)
[-1.200]
81.01
(3)
[-1.253]
85.29
(3)
[-1.402]
18.99
(3)
[1.253]
3.10
(3)
[1.331]
80.99
(3)
[-1.386]
71.51
(3)
[-1.253]
5.64
(3)
[1.221]
82.03
(3)
[-1.395]
16.19
(2)
[-0.481]
8.86
(2)
[-0.684]
28.25
(3)
[0.812]
20.92
(2)
[0.455]
81.01
(3)
[-1.253]
75.13
(3)
[-0.737]
75.51
(3)
[-0.820]
13.40
(2)
[0.629]
80.64
(3)
[-1.400]
5.57
(2)
[-0.595]
18.69
(2)
[-0.612]
79.32
(2)
[0.475]
9.23
(3)
[0.813]
73.16
(2)
[-0.293]
67.12
(2)
[-0.078]
78.99
(2)
[-0.119]
79.31
(2)
[-0.055]
79.31
(3)
[-1.084]
79.98
(1)
[1.065]
5.90
(1)
[-1.253]
 3.  felipecalderero's GSRM sup. KL area-weighted  [normal] S(0.5) 0 0 1 0  82.29 
 (2.31) 
 [-0.429] 
63.92
(1)
[0.873]
10.76
(1)
[-0.761]
4.81
(1)
[-1.049]
15.93
(2)
[0.121]
16.32
(2)
[0.127]
11.25
(3)
[1.161]
11.71
(3)
[0.932]
74.64
(2)
[-0.048]
81.95
(2)
[0.058]
87.65
(1)
[0.862]
18.05
(2)
[-0.058]
2.70
(2)
[-0.252]
82.70
(2)
[0.448]
72.93
(2)
[0.058]
5.29
(2)
[0.007]
83.72
(2)
[0.498]
17.64
(3)
[1.392]
10.85
(3)
[1.414]
28.08
(2)
[0.597]
21.29
(3)
[0.932]
81.95
(2)
[0.058]
75.18
(2)
[-0.676]
75.70
(2)
[-0.587]
13.48
(3)
[0.782]
81.64
(2)
[0.525]
6.08
(3)
[1.409]
20.39
(3)
[1.410]
77.73
(3)
[-1.391]
9.16
(2)
[0.596]
72.44
(3)
[-1.052]
65.87
(3)
[-1.184]
78.19
(3)
[-1.161]
78.46
(3)
[-1.196]
80.24
(2)
[-0.245]
77.22
(3)
[-1.338]
5.92
(3)
[1.194]