Neuronal Structure Segmentation on EM segmentation challenge ISBI 2012 (test)
0.9599Rand Error (Thin)RotEqNet
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| RotEqNetN=2, # params.=30k2016.12 | 0.9599 | — | — | — | 0.9806 | |
| DIVE# params.=5.7M2016.12 | 0.9685 | — | — | — | 0.9858 | |
| PolyMtl# params.=11M2016.12 | 0.9689 | — | — | — | 0.9861 | |
| RotEqNetmode=3 models, # params.=100k2016.12 | 0.9712 | — | — | — | 0.9865 | |
| U-Net# params.=33M2016.12 | 0.9728 | — | — | — | 0.9866 | |
| CUMedVision2016.12 | 0.9768 | — | — | — | 0.9886 | |
| IAL MC/LMC2016.12 | 0.9826 | — | — | — | 0.9894 | |
| DIVE2015.05 | — | 0.0004 | 0.0545 | 0.0582 | — | |
| DIVE-SCI2015.05 | — | 0.0004 | 0.0305 | 0.0584 | — | |
| IDSIANotes=sliding-window convolutional network2015.05 | — | 0.0004 | 0.0504 | 0.0613 | — | |
| IDSIA-SCI2015.05 | — | 0.0007 | 0.0189 | 0.1027 | — | |
| u-netAugmentation=averaged over 7 rotated versions of input data, Post-processing=none2015.05 | — | 0.0004 | 0.0382 | 0.0611 | — |