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 (deepv)
set \ criterionCSOSUSMENEOCCACOCCIIIEAMSRMCIGCELCEBCEGBCEBGMSCSSCVDLAVINVINMIMARIJCDCFMIWIWIINBDE
#3 - all99.940.000.000.000.0099.9899.980.020.030.0499.9735.130.03-31.843.340.030.120.100.140.1299.9499.8899.880.0699.940.080.4899.520.0899.8299.7699.8899.8899.8899.880.04
#4 - all99.380.000.000.000.0099.8899.910.050.110.1199.8928.480.11-37.268.350.111.240.951.521.2499.3898.7698.760.6299.330.652.9497.000.6098.4697.7798.8798.8798.8698.880.38
#5 - all74.100.000.0025.8125.7399.9699.950.030.080.0599.9215.240.06-37.3210.540.068.324.3412.688.7092.5988.5088.297.4188.552.098.1091.182.8592.5089.4294.4294.4294.0894.751.60
#6 - all98.540.000.000.000.0099.9599.950.100.170.7299.8313.480.21-44.6420.970.302.852.383.352.8898.5497.1497.131.4698.161.334.6494.691.2196.7395.2097.5497.5497.1797.910.91
#7 - all98.860.000.000.000.00100.00100.000.110.170.3599.8314.080.23-44.269.740.242.201.512.932.2398.8697.7997.761.1498.511.033.3296.460.6197.8996.5898.2698.2698.0898.440.49
#8 - all99.540.000.000.000.00100.00100.000.050.100.0999.9014.080.09-42.615.810.090.920.791.060.9299.5499.0899.080.4699.430.541.6298.320.2299.0998.4699.2299.2299.2299.230.13
#9 - all61.790.000.0037.5834.52100.00100.005.427.2712.1792.7310.799.09-34.929.139.4014.397.9722.7916.3688.8481.7081.8711.1689.804.4712.7085.626.1877.6668.5581.3481.5687.8175.769.48
#10 - all98.760.000.000.000.0099.9999.990.160.280.4099.7210.360.32-43.534.630.332.452.112.792.4698.7697.5597.561.2498.531.283.4696.440.4897.6295.8697.8997.8998.0397.750.69
#11 - all99.030.000.000.000.00100.00100.000.070.110.2099.897.900.13-39.776.300.141.881.532.261.9199.0398.1098.120.9798.600.932.4397.450.3198.4697.3198.6498.6498.7298.560.33
#12 - all96.960.000.001.031.38100.00100.000.010.020.0399.986.890.02-44.909.700.023.912.954.873.9297.9196.0796.032.0995.631.754.3995.130.6597.0294.9297.3997.4097.0397.760.90
#6 - bark97.970.000.000.000.00100.00100.000.150.490.2399.5112.200.29-45.7615.380.323.963.494.433.9697.9796.0696.062.0397.091.816.3093.091.3396.1094.1096.9696.9696.8897.041.00
#6 - flowers73.880.0023.580.001.2699.6499.699.3915.359.5984.6512.1411.74-22.6213.8912.104.643.3214.8813.5589.8788.0285.716.3486.862.779.6389.383.7289.1684.3891.5391.6986.3497.382.47
#6 - glass97.140.000.000.000.0099.9399.790.120.330.2999.6715.690.23-40.5814.400.275.244.096.715.5697.1494.6894.492.8695.922.267.8891.512.0293.8790.7595.1595.1693.8796.471.68
#6 - man-made99.240.000.000.000.00100.00100.000.170.310.3599.6919.510.33-35.195.070.331.511.201.821.5199.2498.4998.490.7699.310.792.7797.050.6797.9496.7798.3698.3698.3998.330.60
#6 - nature99.670.000.000.000.00100.00100.000.010.020.0399.9813.470.02-20.8122.250.020.650.610.700.6599.6799.3599.350.3399.390.391.3798.070.3199.3699.2299.6199.6199.6399.590.21
#6 - plants95.980.000.000.000.0099.8699.860.360.590.8899.4116.830.71-35.014.690.726.864.949.097.1695.9892.6992.504.0294.952.478.5991.122.2592.5988.6493.9893.9893.4694.505.50
#6 - rock98.530.000.000.000.0099.8399.780.150.290.3499.7119.550.29-41.409.120.302.912.423.392.9198.5397.1097.101.4798.351.394.8594.891.1096.5194.5497.1997.1996.9897.410.57
#6 - stone99.570.000.000.000.00100.00100.000.010.020.0299.9811.150.01-47.6529.200.010.850.671.040.8599.5799.1599.150.4398.810.461.6297.770.2999.3799.1799.5999.5999.6499.531.26
#6 - textile99.620.000.000.000.0099.9399.840.050.100.1699.9019.750.10-40.1711.000.120.750.670.840.7599.6299.2599.250.3899.470.451.5798.300.2699.2298.7899.3899.3899.3499.430.13
#6 - wood95.160.000.003.951.89100.00100.000.010.010.0699.9913.770.01-39.7419.550.024.423.446.695.7297.0495.2294.242.9692.352.006.9791.961.9494.7792.4096.0596.0794.3097.875.68
mean94.180.001.183.423.2499.9599.940.821.291.3098.7115.521.20-38.5011.651.253.502.475.204.1797.4095.7395.542.4196.451.454.7894.751.3595.7193.6396.5696.5896.3996.821.70
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