-
你好，有一些问题需要咨询一下
1. 请问出现了梯度无法回传的问题，是如何解决的，根据本人的调试，需要将RCE擦除策略中设置x2并且令 x2=x.clone()，最后返回x2可以解决。
2.为什么模型训练2次？在train.py的main函数中model.train,同时在train函数中也有model.train
3.为什么训练时候一直不收敛，loss收敛到0.5左右不再变化
希望能得到回…
-
Greetings Sir, I would like to know the title of the paper where EDAE-Net is from. thank you
-
Hello author, the content of your paper is very wonderful!!!
I want to try to run the code but I don't know where to find the code, can you help me?
-
Hi,
Thanks for the great work and the released code! I am very interested in your work and have tried to reimplement the work and results. During the reimplementation, I encountered an issue and ho…
-
Dear author:
The architecture is very powerful for both vessel segmentation and A/V classification, yet I have a question about the result metrics, could you explain why the segmentation use AUC…
-
For the moment, here is the typical pipeline to obtain the blood flow in the retina's blood vessels :
1. Use the real-time visualisation of the eye with Holovibes to get a good focus on the eye of…
-
I plan to create a demo that can be used for liver segmentation prediction. How can I extract the code for use in my own program
-
Hi carlesventura,
I tried to run the code vessels/iterative/iterative_graph_creation_no_mask_offset.py, and it reports an error: cannot find the results_dir = './results_dir_vessels/results_DRIU_vess…
-
Hi,
I am using clodsa to achieve data augmentation for semantic segmentation. And now I meet the error "None-Type". I am trying the debug in pycharm and found the error point in `folderLinearSemant…
-
Hi all,
Could anyone help in this issue. Why when i use this cldice loss for training my vessel segmentation the loss is all zero throughout the training. What could have happened to this?
![the cl…