MedSegDiff

MedSegDiff copied to clipboard

Where is the SS-Former module in the implementation V2?

21

Hello. I cannot find the cross attention module named SS-Former in the implementation V2. The Generic_Unet returns anchor condition and segmentation only. Can you show me the corresponding position of...

ZhangXiaotong015

Question about the result on other dataset

9

作者您好，我有以下几个疑问： 1.在gaussian_diffusion/unet.py中UNetModel_newpreview表示v2版本，那么与文章对应的SS-Former的部分具体应用在哪里呢？因为我注意到middle_block部分并没有用到self.highway_forward的成分。请问SS-Former对应于代码中的哪一部分？ 2.我利用公开的病理数据做二值分割，训练了400000step后使用--num_ensemble 5和--diffusion_steps 1000，却得到以下结果： output_ens.jpg  ens0.jpg  ens1.jpg ens2.jpg ens3.jpg ens4.jpg 请问这是为什么？是有什么需要特别注意的地方吗？ 3.当使用--dpm_solver True 的时候，合成的结果却是这样的：  请问一下产生这种图像的话应该如何解决？ 期待您的回复！感谢！

JikC

RuntimeError: a leaf Variable that requires grad is being used in an in-place operation.

6

Traceback (most recent call last): File "segmentation_train.py", line 117, in main() File "segmentation_train.py", line 69, in main TrainLoop( File "E:\MedSegDiff-master2.0\guided_diffusion\train_util.py", line 83, in __init__ self._load_and_sync_parameters() File "E:\MedSegDiff-master2.0\guided_diffusion\train_util.py", line 139, in...

DWBSIC

infer

2

Hello, may I ask that the shape of your inference code output is [b,2,w,h]. Why only use the -1 dimension

1390806607

"cal" variable usage

2

Thank you for your excellent work！ I have a question about condition encoding branch: Additionally to _out_ (~_uemb_ later) variable, _cal_ variable is returned. As I understand, cal was used...

ser5kovskiy

res = torch.where(mask > 0, 1, 0) #merge all tumor classes into one to get a binary segmentation mask Hello, I don't quite understand why all categories of masks are...

1390806607

how to get the best dice checkpoint during training?

7

Good job！ But how can i get the best dice checkpoint during training? It seem there is no relative eval code in the project during training.

LiuTingWed

How to run multi-label segmentation?

15

I encountered some issues with multi-label segmentation, and I would like to ask for your help. The demo ISIC dataset has a single label and the demo BRATS dataset has...

gulubao

About the image size

1

When I trained with the previous version, the code worked fine when the dataset size was 256, but I got an error when my dataset size was 512.How should I...

kjRainy

error in sample

3

First of all, thanks for your excellent work! I met a problem that when I run `segmentation_sample.py`, it only sample and infer 2 images. I wonder how can I solve...

wzx0720