前言

接下來我們來討論一下比較的內容。

Rule

We demonstrate the application of the u-net to three different segmentation tasks.

因為本篇的主題是Unet所以他在不同的task上面做測試。

The first task is the segmentation of neuronal structures in electron microscopic recordings. An example of the data set and our obtained segmentation is displayed in Figure 2. We provide the full result as Supplementary Material.

第一個任務為電子顯微鏡的細胞分割。

The data set is provided by the EM segmentation challenge [14] that was started at ISBI 2012 and is still open for new contributions. The training data is a set of 30 images (512x512 pixels) from serial section transmission electron microscopy of the Drosophila first instar larva ventral nerve cord (VNC). Each image comes with a corresponding fully annotated ground truth segmentation map for cells (white) and membranes (black).

敘述資料級與比賽的細節
(1) 照片大小
(2) 照片數量
(3) 分類類別

The test set is publicly available, but its segmentation maps are kept secret.

測試資料保密

An evaluation can be obtained by sending the predicted membrane probability map to the organizers. The evaluation is done by thresholding the map at 10 different levels and computation of the “warping error”, the “Rand error” and the “pixel error” [14].

並說明了評分的標準
包含了三種誤差評估方式：warping error, Rand error, and pixel error
但是看了官方網站，目前Pixel error 與 rand error 已經被棄用。

Concusion

所以參加比賽之前最好還是看一下評估的準則。

Reference

[0] ISBI Challenge: Segmentation of neuronal structures in EM stacks