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 (with ewt)
set \ criterionCSOSUSMENEOCCACOCCIIIEAMSRMCIGCELCEBCEGBCEBGMSCSSCVDLAVINVINMIMARIJCDCFMIWIWIINBDE
#3 - all99.960.000.000.000.00100.00100.000.010.020.0399.9835.130.02-31.853.340.020.070.060.080.0799.9699.9399.930.0499.970.050.2899.710.0599.8999.8599.9299.9299.9299.920.02
#4 - all99.600.000.000.000.0099.9199.890.060.130.1299.8728.580.12-37.348.290.120.800.720.870.8099.6099.2199.210.4099.600.462.0697.890.4298.9398.4599.2299.2299.2199.220.20
#5 - all99.500.000.000.000.0099.8899.910.120.200.3599.8013.720.23-37.1912.130.250.990.881.100.9999.5099.0199.010.5099.230.532.0597.770.3199.1898.7899.3999.3999.4499.330.23
#6 - all98.870.000.000.000.0099.99100.000.100.140.6699.8613.720.21-44.7820.780.282.231.752.722.2498.8797.7797.771.1398.701.113.8795.571.0097.2996.0097.9697.9697.9897.931.03
#7 - all98.950.000.000.000.00100.00100.000.010.020.0699.9814.030.02-44.539.830.031.981.172.822.0198.9597.9797.951.0598.730.762.4597.390.5598.0996.8998.4298.4298.2698.580.52
#8 - all99.200.000.000.000.00100.00100.000.060.130.1199.8714.060.11-42.595.820.111.591.371.811.5999.2098.4198.420.8099.040.832.5097.400.3998.4297.3398.6598.6598.7798.530.25
#9 - all78.1228.040.0015.5213.5499.9999.990.060.080.2099.9211.830.11-45.3810.620.1311.726.7420.9816.0187.4181.9483.098.6774.624.3912.4585.934.1485.2078.0587.6787.8393.2982.698.44
#10 - all98.210.000.000.000.0099.9599.940.280.460.6799.5410.330.54-43.264.550.553.472.654.323.5098.2196.5096.521.7997.921.624.3995.490.7196.4893.9496.8796.8897.3896.380.53
#11 - all98.920.000.000.000.0099.9899.970.080.120.2799.887.920.17-39.796.260.182.111.632.612.1398.9297.8797.891.0898.561.022.6597.220.3698.2196.8898.4198.4198.6898.150.34
#12 - all97.057.494.690.000.00100.00100.000.080.130.1899.876.810.15-44.729.640.153.921.945.953.9797.0595.3495.332.1591.431.523.8195.780.7396.6594.3197.0797.0797.5896.560.97
#6 - bark98.500.000.000.000.00100.00100.000.030.070.0699.9312.420.06-46.3815.020.062.891.963.882.9598.5097.0597.091.5098.031.324.6094.971.0796.8695.2097.5497.5498.3696.730.71
#6 - flowers99.430.000.000.000.0099.9799.970.030.070.0699.9313.030.05-42.4113.130.061.120.991.261.1299.4398.8898.880.5799.260.622.1597.710.3498.9698.3699.1799.1799.2399.120.17
#6 - glass98.610.000.000.000.0099.9799.940.080.120.4299.8815.900.16-40.8113.970.212.732.303.192.7698.6197.2797.251.3998.321.314.5695.111.0796.7294.9397.4097.4097.0997.710.82
#6 - man-made99.280.000.000.000.00100.00100.000.010.030.0399.9719.530.03-35.625.150.031.421.091.761.4399.2898.5798.580.7299.230.762.6397.200.5498.3497.3998.6898.6898.9398.430.32
#6 - nature99.660.000.000.000.00100.00100.000.010.010.0299.9913.480.02-20.9122.270.020.680.430.940.6999.6699.3199.320.3499.510.381.3298.140.4099.1698.9899.4899.4999.7399.240.15
#6 - plants99.360.000.000.000.0099.9999.980.010.030.0399.9716.970.03-35.914.540.031.271.141.391.2799.3698.7498.740.6499.360.702.4397.500.4698.4997.5798.7798.7798.7598.790.37
#6 - rock98.670.000.000.000.0099.7699.810.220.400.5299.6019.570.43-41.318.980.452.622.282.982.6398.6797.3797.381.3398.571.244.3395.441.0196.7794.9397.4097.4097.4897.320.55
#6 - stone99.800.000.000.000.00100.00100.000.010.010.0599.9911.280.02-47.7129.140.030.390.270.530.4099.8099.6099.600.2099.480.220.7798.940.1599.6699.5599.7899.7899.9299.630.09
#6 - textile99.710.000.000.000.0099.9899.980.050.070.2499.9319.780.11-40.2310.950.130.580.480.670.5899.7199.4299.420.2999.640.341.1998.710.2399.3298.9399.4699.4699.5299.400.12
#6 - wood98.170.000.000.000.00100.00100.000.030.040.2199.9613.570.07-39.7019.240.093.322.294.463.4398.1796.5496.501.8396.481.304.5194.850.8997.5796.3898.1598.1597.9698.351.06
mean97.981.780.230.780.6899.9799.970.070.110.2199.8915.580.13-40.1211.680.152.291.613.222.5398.4497.3397.391.3297.281.023.2596.440.7497.5196.1397.9797.9898.3797.600.85
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