Diff-UNet icon indicating copy to clipboard operation
Diff-UNet copied to clipboard

trained Model

Open maoerlu opened this issue 1 year ago • 12 comments

您好,能提供一下您训练好的模型吗?我的邮箱是[email protected],万分感谢

maoerlu avatar Apr 19 '23 10:04 maoerlu

I will send you the segmentation model on brats2020 dataset when I have downloaded it from the server.

920232796 avatar Apr 20 '23 07:04 920232796

I will send you the segmentation model on brats2020 dataset when I have downloaded it from the server.

thanks. I have tested the model you sent me, but I have not received the results from your paper.do you have any idea why is that? Is the test set different? do you have a better model?

My results are as follows: wt =0.9325 tc = 0.8292 et = 0.6654 wt_hd =1.0293 tc_hd = 4.0204 et_hd = 6.0463 wt_recall = 0.9279 tc_recall = 0.8033 et_recall = 0.7266

maoerlu avatar Apr 29 '23 12:04 maoerlu

I will check it and get back to you later. Thank you for your attention.

920232796 avatar May 02 '23 22:05 920232796

I will check it and get back to you later. Thank you for your attention. thank you.

maoerlu avatar May 03 '23 06:05 maoerlu

I have run the testing code again using the model I send you and I get this result when I set the S=1.

image

The dice of WT, TC and ET is 0.9198, 0.8665, 0.7564, respectively.

920232796 avatar May 03 '23 16:05 920232796

作者您好,我租不到太多GPU,能提供一下您已经训练好的模型吗?邮箱:[email protected]

XFivezzz avatar May 07 '23 02:05 XFivezzz

作者您好,我没有太好的GPU进行训练, 能提供一下您已经训练好的模型吗?邮箱:[email protected]

Ligo04 avatar Jun 15 '23 09:06 Ligo04

作者您好,我没有太好的GPU进行训练, 能提供一下您已经训练好的模型吗?邮箱:[email protected],万分感谢!

Benbie avatar Sep 21 '23 15:09 Benbie

作者您好,能提供一下已经训练好的模型吗?邮箱:[email protected],万分感谢!

JoeQvQ avatar Oct 11 '23 03:10 JoeQvQ

作者您好,能提供一下已经训练好的模型吗?邮箱:[email protected],万分感谢!

SillenceHu avatar Nov 08 '23 07:11 SillenceHu

作者您好,我没有太好的GPU进行训练, 能提供一下您已经训练好的模型吗?邮箱:[email protected]

ZhengChen6 avatar Mar 24 '24 08:03 ZhengChen6

链接: https://pan.baidu.com/s/1zCdMeJKQ9KCVQVPoSaSZOQ?pwd=31pw 提取码: 31pw --来自百度网盘超级会员v8的分享

ge-xing avatar Mar 24 '24 08:03 ge-xing