The Prague Texture Segmentation Datagenerator and Benchmark - Result Details
optimized
for

It is possible that the resulting index maps don't correspond to the training set's indeces according to the data.xml file!
It leads to the wrong values of pixel-wise criteria as a result.


Benchmark - Colour [normal] CLresult byhuangyuan's unet (psp21)
set \ criterionCSOSUSMENEOCCACOCCIIIEAMSRMCIGCELCEBCEGBCEBGMSCSSCVDLAVINVINMIMARIJCDCFMIWIWIINBDE
#3 - all99.860.000.000.000.000.130.1299.7299.8699.860.140.0799.8699.790.0299.860.280.250.310.2899.8699.7299.720.1499.860.181.0598.950.1999.5799.4399.7299.7299.7199.720.09
#4 - all98.570.000.000.000.001.501.2897.1998.5798.581.430.4798.5797.860.3098.572.822.323.332.8298.5797.1897.191.4398.491.366.1293.751.4196.3894.8197.3497.3497.3797.300.90
#5 - all99.350.000.000.000.000.940.5798.7299.3599.350.650.1599.3599.030.0599.351.291.181.391.2999.3598.7298.720.6599.030.672.5997.180.4398.8898.3499.1699.1699.1399.190.31
#6 - all97.700.000.000.000.001.721.0695.5197.7097.752.300.9997.7096.550.6997.714.453.045.884.4797.7095.5195.512.3097.711.876.5092.562.3193.7891.0695.3295.3294.8295.821.04
#7 - all97.010.000.000.000.002.092.7894.4697.0197.172.990.4197.0395.510.5897.065.434.396.635.5997.0194.4194.462.9995.152.267.2692.281.3695.2492.4296.0696.0696.1595.981.10
#8 - all98.280.000.000.000.001.702.1496.6498.2898.291.720.2898.2897.430.1698.293.362.853.883.3898.2896.6496.641.7297.981.594.7695.040.8696.5594.2797.0597.0596.8297.280.63
#9 - all83.830.000.0010.2713.203.745.8287.6092.9193.347.090.8192.7189.361.8392.9211.277.0015.6711.4092.9187.7487.686.8283.413.7310.6087.753.7087.0780.6789.3089.3491.9786.796.35
#10 - all95.300.000.000.000.003.954.6391.0695.3095.454.700.5995.2892.940.6295.338.896.6111.238.9595.3091.1091.064.7094.903.609.7689.931.9090.6284.6691.6991.7091.1292.281.68
#11 - all97.280.000.000.000.001.552.5594.7597.2897.342.720.2697.2795.920.3097.295.184.016.435.2697.2894.8094.752.7296.462.275.9093.790.9795.2191.8695.7595.7595.9895.530.82
#12 - all95.753.338.240.000.003.373.4689.7393.2295.316.780.4592.8389.831.1993.366.234.359.147.2595.7592.9692.373.5988.652.736.8692.361.3094.0990.1894.8494.8494.2695.410.95
#6 - bark98.360.000.000.000.001.371.9796.7998.3698.361.640.4098.3697.540.1798.363.202.763.663.2298.3696.8096.791.6497.831.515.2794.221.1796.5794.7997.3397.3397.2497.410.70
#6 - flowers87.450.000.0010.427.6118.962.8770.4171.2775.1328.733.7172.9656.914.5973.088.545.5212.119.0994.8390.7690.565.1792.973.0910.7588.472.7891.5887.4993.3393.3392.6294.052.35
#6 - glass95.890.000.000.000.003.604.4592.2095.8995.884.111.1595.8493.840.5995.867.606.369.127.8895.8992.3892.204.1194.503.0310.5488.662.9191.1686.9193.0093.0091.9994.022.13
#6 - man-made98.850.000.000.000.001.041.3297.7398.8598.851.150.3098.8598.280.1398.852.281.942.612.2898.8597.7397.731.1598.801.184.1195.620.9097.2295.6797.7997.7997.7997.790.56
#6 - nature98.790.000.000.000.001.961.4297.6298.7998.791.210.7298.7998.180.1398.792.382.052.732.3998.7997.6297.621.2197.801.174.0894.231.1997.5097.0098.4898.4898.3098.650.64
#6 - plants95.830.000.000.000.001.733.4392.0695.8395.994.170.9095.7993.741.0095.857.635.579.897.8295.8392.1992.064.1795.152.9310.2189.462.5891.5187.0993.1093.1092.5693.654.84
#6 - rock97.300.000.000.000.003.193.0194.7697.3097.332.700.6997.3095.950.3297.315.264.476.055.2797.3094.7594.762.7097.172.247.8191.772.0793.4189.9394.7094.7094.6194.781.24
#6 - stone98.870.000.000.000.001.312.5597.8098.8798.881.130.2798.8798.300.1798.872.211.812.632.2398.8797.7997.801.1397.171.073.7494.840.8998.0297.4398.7098.7098.9598.450.62
#6 - textile99.360.000.000.000.000.550.7098.7399.3699.360.640.1799.3699.040.0499.361.281.181.371.2899.3698.7398.730.6499.150.722.5097.280.4798.5897.7998.8898.8898.8198.950.27
#6 - wood95.430.000.003.951.680.551.6694.3597.1197.202.891.0496.6795.661.1296.903.822.676.715.5797.1195.5594.352.8991.781.766.1392.911.8095.1592.9496.3496.3794.2098.581.08
mean96.450.170.411.231.122.752.3993.8996.0696.413.940.6996.0894.080.7096.154.673.526.044.8997.3695.1595.032.5995.701.956.3393.051.5694.9092.2495.8995.9095.7296.081.42
criterionGraph
correct
over
under
missed
noise
F-measure
imageTexture mosaic (undegraded)Ground truthSegmentation
#3 - all
#4 - all
#5 - all
#6 - all
#7 - all
#8 - all
#9 - all
#10 - all
#11 - all
#12 - all
#6 - bark
#6 - flowers
#6 - glass
#6 - man-made
#6 - nature
#6 - plants
#6 - rock
#6 - stone
#6 - textile
#6 - wood