Eight3H
Eight3H
> I ran train.py and inference.py according to your method, and after running inference.py, a pred folder was generated, which contains all the split results of the validation set, but...
> > Hi, > > I am not the expert of this code, but maybe I can help, until @XuJiacong gets more detailed feedback. Noticed that the load checkpoint process...
你好，你现在有吗
> 我在验证集上预测之后去官网上进行评估，dice的结果如下TC: 0.71826, WT: 0.90351, ET: 0.8189，为什么效果会差这么多？尤其是在TC和ET的dice 提交到官网评估的是分割后的什么
> 我在验证集上预测之后去官网上进行评估，dice的结果如下TC: 0.71826, WT: 0.90351, ET: 0.8189，为什么效果会差这么多？尤其是在TC和ET的dice 你好，可以提供一下联系方式吗
没成功，生成的图片有问题，颜色很奇怪
> Hello, how do you get the json file when you validate? Can you send me the contents of the json file?Thank you very much Generate on your own, based...
> Hello, how do you get the json file when you validate? Can you send me the contents of the json file?Thank you very much Have you submitted it to...
> Hello, I also encountered the same issue during training, with particularly low accuracy. Have you managed to resolve it? @FengheTan9 Leave a contact information and discuss it
> 我跟原文一样的设置，验证时候ET总是特别低 你用的是几几年的数据集