happy20200

Results 23 comments of happy20200

你好,我现在也是这个问题,训练过程中的测试各个指标都是0,但predict.py效果还行,这是怎么回事啊

Where can I find the test file? Why can't I find it? @wrbbb

@manvirvirk May I ask if you solved this problem, I'm experiencing the same problem now!

> 在使用提供的图像数据进行训练和测试时,我可以产生良好的结果,但是当转到我自己的数据集时,即带有肺罩的胸部X射线图像,测试结果图像基本上都是黑色的,有时在边缘。训练期间的指标“准确性”高于0.9,但输出图像很差,这就是我感到困惑的地方。我尝试用骰子替换指标,但结果仍然是黑色的。 May I ask if you solved this problem, I was testing directly with my own data but it didn't work well. ![image](https://user-images.githubusercontent.com/59427046/84889860-25aaf800-b0cc-11ea-8a1f-b3c1aa0c0ea2.png)

Where can I find the test file? Why can't I find it? @n12iki

Where can I find the test file? Why can't I find it? @mfaramarzi

> 我通过在model.py中 > 删除一行代码解决了该问题:删除此代码: > conv9 = Conv2D(2,3,activation ='relu',padding ='same',kernel_initializer ='he_normal')(conv9) > 请问测试文件在哪呢,我怎么都没有找到

> > 导入错误:无法从“assets.ops.dcn”导入名称“deform_conv_cuda” > > **大佬,解决了吗?你能告诉我吗?** 请问最后解决了吗

> Hi @Tian14267 I faced with the same issue. You need to "build deformable convolution opertor" as mentioned in the README > > > build deformable convolution opertor > >...