ericelmoznino / deep_bilateral_network Goto Github PK

Sorry, I have sone problem installing setup.py. I dont't know why?

bilateral_slice.cpp(85): error C2059: 语法错误:“do”
bilateral_slice.cpp(88): error C2143: 语法错误: 缺少“;”(在“{”的前面)
bilateral_slice.cpp(88): error C2447: “{”: 缺少函数标题(是否是老式的形式表?)
bilateral_slice.cpp(85): error C2059: 语法错误:“while”
bilateral_slice.cpp(89): error C2059: 语法错误:“do”
bilateral_slice.cpp(89): error C2143: 语法错误: 缺少“;”(在“{”的前面)
bilateral_slice.cpp(89): error C2447: “{”: 缺少函数标题(是否是老式的形式表?)
bilateral_slice.cpp(89): error C2059: 语法错误:“while”
bilateral_slice.cpp(92): error C2059: 语法错误:“if”
bilateral_slice.cpp(92): error C2143: 语法错误: 缺少“;”(在“{”的前面)
bilateral_slice.cpp(92): error C2447: “{”: 缺少函数标题(是否是老式的形式表?)
bilateral_slice.cpp(97): error C2059: 语法错误:“else”
bilateral_slice.cpp(97): error C2143: 语法错误: 缺少“;”(在“{”的前面)
bilateral_slice.cpp(97): error C2447: “{”: 缺少函数标题(是否是老式的形式表?)
bilateral_slice.cpp(104): error C2065: “bs”: 未声明的标识符
bilateral_slice.cpp(104): error C2065: “h”: 未声明的标识符
bilateral_slice.cpp(104): error C2065: “w”: 未声明的标识符
bilateral_slice.cpp(105): error C2065: “input”: 未声明的标识符
bilateral_slice.cpp(108): error C2059: 语法错误:“return”
bilateral_slice.cpp(122): error C2059: 语法错误:“}”
bilateral_slice.cpp(122): error C2143: 语法错误: 缺少“;”(在“}”的前面)
bilateral_slice.cpp(130): error C2143: 语法错误: 缺少“;”(在“{”的前面)
bilateral_slice.cpp(130): error C2447: “{”: 缺少函数标题(是否是老式的形式表?)
bilateral_slice.cpp(182): error C2065: “bilateral_slice_backward”: 未声明的标识符

Hello, thank you for your work! I have prepared the dataset and then run the code. When I train the model Curves or NN use "python train.py Curves" there exists an error like this and I don't know how to solve it so I need your help. Looking forward for your reply~ Thanks.
Starting epoch: 1 Segmentation fault (core dumped)

能参考一下数据集是什么样的么？

I have got the model running and the result seems good now.
The question is how to choose the training method" NN and curve", what's pros and cons?

By the way, I am now finding ways to port to coreml and ncnn.

Thank you

what is the parameter setting in configuration.py for best/model.pth?

Hi, Thanks for your wonderful work.
I'm using the pre-trained model in saved_runs/best/model.pth to test my own images,
and some strange artifacts are shown in the squared area as below:

Could you please give me any advice on finding the bug?

Have you been able to implement the bilateral slicing for coreml in ios?