3DUX-Net icon indicating copy to clipboard operation
3DUX-Net copied to clipboard

Results 32 3DUX-Net issues
Sort by recently updated
recently updated
newest added

Hello, thank you for your work. I read that you used a total of 511 flare 2021 datasets in the article, but I only saw 361 data and 50 unlabeled...

Hello, I wanna train with AMOS2022 dataset, but I don't know what I need to config firstly. I find the TRAINING.md, in which there are no "INSTALL.md" and AMOS2022 Finetuning...

![螢幕擷取畫面 2024-03-15 005435](https://github.com/MASILab/3DUX-Net/assets/158141789/6b16095c-27ad-4b09-87fa-21717db8d8aa) how can i solve this problem I want to run my own data. Which part of the code do I need to modify? Best regard, Andrew

我微调了一下代码,让模型用两个显卡跑,还是会出现显卡内存不足情况 。 把 max_iter 设为10万,eval_step设为1000,batch_size设为1,num_workers 设为0,用两张显卡跑也出现了显卡内存不足情况,但明明显示两个显卡内存分别占用7G左右,显卡是有充足的内存,不知什么原因了,请教您一下各位大佬解决问题的方法 ,谢谢! ![image](https://github.com/MASILab/3DUX-Net/assets/139946751/74be4a8a-abd2-4f16-8c95-17c9e765c362)

The 3 Json file just mentioned training and testing sets, but nothing about validation sets. Is validation sets was used in the project? or is it necessry? I saw in...

Hi, I got error when using SWINUNETR, my input size to the model is 4,1,96,96,96. When changing to 3DUXNET it works. > raise ValueError("Expected more than 1 spatial element when...

![image](https://github.com/MASILab/3DUX-Net/assets/72400883/e1e12e50-264d-4e4c-aef9-3258dd62c941) Hello, I used your model to train on Flare2021, but the inference result seems to be missing something (Kidney miss). Can you help me solve the problem

Dear Author, is there any difference between the code "main_finetune" and "main_train"? What does it do?I look forward to your reply and thank you very much.

Your figure is not consistent to your code. There are 5 stages in your figure while only 4 stages in your code. Futhermore, the downsampling should be ahead of UX-Net...

Hi, I would like to ask about using my own dataset. I have volume data with size 200,200,100 (D,W,H) and corresponding label with same size. I don't have metadata such...