Skip to main content

Table 6 Segmentation performance for the ATUM-SEM dataset

From: Effective automated pipeline for 3D reconstruction of synapses based on deep learning

Size of pixel Metrics Sun [18] U-Net [23] F-RCNN+ F-RCNN + F-RCNN + fusion +
overlap     Growing [19] GrabCut continuity + GrabCut
0-pixel AP 32.7% 55.4% 49.0% 57.6% 65.6%
  F1 36.3% 58.1% 53.4% 60.6% 65.2%
  Jaccard 30.6% 49.5% 43.2% 52.6% 58.1%
3-pixel AP 56.3% 67.4% 67.8% 78.8% 86.1%
  F1 60.6% 66.9% 69.2% 75.8% 81.5%
  Jaccard 50.1% 59.0% 59.8% 69.0% 74.7%
5-pixel AP 59.4% 70.6% 76.4% 80.3% 88.6%
  F1 62.7% 70.1% 74.4% 77.5% 84.1%
  Jaccard 53.7% 60.2% 63.5% 72.2% 76.9%