Chenyang's Notes

This is Chenyang's notes for issue #120.

---

inference with tutorial image:

inference with validation image:

inference with test image:

---

2D (xy-axis) version of groundtruth (z 0~249):

z=10

z=125

z=240

2D (xy-axis) version of inference results (z 0~249):

z=10

z=125

z=240

2D (xy-axis) version of provided results (z 0~249):

z=10

z=125

z=240

---

Data used: neuroproof_examples

(i) Training data: size 520*520*520 (with groundtruth)

(ii) Testing data: size 250*250*250

The settings of Flood-filling Networks:

The model:

The inference result:

(i) The single segmentation, starting from the specified origin point (zy-axis, yx-axis, xz-axis):

(ii) The segmentation from default function:

(iii) The segmentation from seg-0_0_0.npz: