Microtubule Segmentation on MicSim FluoMT-Complex synthetic (test)
0.017LossMTFlow
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| MTFlowBackbone=Four down-sampling and up-sampling blocks, Number of filters=642026.01 | 0.017 | 82.28 | 78.67 | 86.24 | 0.8169 | 91.47 | |
| ResUnetBackbone=Four down-sampling and up-sampling blocks, Number of filters=642026.01 | 0.0194 | 76.72 | 90.67 | 66.49 | 0.7663 | 89.74 | |
| U-Net++Backbone=Four down-sampling and up-sampling blocks, Number of filters=642026.01 | 0.0195 | 77.88 | 89.42 | 68.94 | 0.7755 | 90.07 | |
| U-NetBackbone=Four down-sampling and up-sampling blocks, Number of filters=642026.01 | 0.0202 | 76.81 | 89.53 | 67.25 | 0.7658 | 89.39 | |
| TransUNetBackbone=Four down-sampling and up-sampling blocks, Number of filters=162026.01 | 0.0309 | 63.89 | 85.28 | 51.07 | 0.6432 | 72.58 | |
| CAR-UNetBackbone=Four down-sampling and up-sampling blocks, Number of filters=322026.01 | 0.0472 | 62.78 | 75.32 | 56.21 | 0.6261 | 72.81 |