About mean_std.npy

[Alanzhang92]

Hello, I'm very interest in your work. When I tried to run train.py, it remind me that I need to provide mean_td.npy for both train and test. However, I can't find this file in this project. I think you maybe forget to add it in your project?

[huiqu18]

Sorry for the late reply. The mean_std.py is computed using the images in the training set.

[thinkInJava33]

sorry, Could you please provide the file or the values of mean and std

Sorry for the late reply. The mean_std.py is computed using the images in the training set. sorry, Could you please provide the file

[MegumiOuO]

sorry, Could you please provide the file or the values of mean and std

Sorry for the late reply. The mean_std.py is computed using the images in the training set. sorry, Could you please provide the file

this is the mean_std value calculated by myself，you can use it. normalize: [[0.78780321 0.5120167 0.78493782], [0.16766301 0.24838048 0.13225162]]

[thinkInJava33]

sorry, Could you please provide the file or the values of mean and std

Sorry for the late reply. The mean_std.py is computed using the images in the training set. sorry, Could you please provide the file

this is the mean_std value calculated by myself，you can use it. normalize: [[0.78780321 0.5120167 0.78493782], [0.16766301 0.24838048 0.13225162]]

thank you,but i still got some probleams,i want to train the model on GlaS.But errors occured,like-- FileNotFoundError: [Errno 2] No such file or directory: './data/GlaS/weight_maps/train' do you know how to handle the dataset so that i can run train.py without error.

[MegumiOuO]

sorry, Could you please provide the file or the values of mean and std

Sorry for the late reply. The mean_std.py is computed using the images in the training set. sorry, Could you please provide the file

this is the mean_std value calculated by myself，you can use it. normalize: [[0.78780321 0.5120167 0.78493782], [0.16766301 0.24838048 0.13225162]]

thank you,but i still got some probleams,i want to train the model on GlaS.But errors occured,like-- FileNotFoundError: [Errno 2] No such file or directory: './data/GlaS/weight_maps/train' do you know how to handle the dataset so that i can run train.py without error.

this code could use the weight map to improve the results, and the code for calculating weight map is provided (weight_map.m). You can use this code to calculate it, the input data is anno of the img.

[thinkInJava33]

sorry, Could you please provide the file or the values of mean and std

Sorry for the late reply. The mean_std.py is computed using the images in the training set. sorry, Could you please provide the file

this is the mean_std value calculated by myself，you can use it. normalize: [[0.78780321 0.5120167 0.78493782], [0.16766301 0.24838048 0.13225162]]

thank you,but i still got some probleams,i want to train the model on GlaS.But errors occured,like-- FileNotFoundError: [Errno 2] No such file or directory: './data/GlaS/weight_maps/train' do you know how to handle the dataset so that i can run train.py without error.

this code could use the weight map to improve the results, and the code for calculating weight map is provided (weight_map.m). You can use this code to calculate it, the input data is anno of the img.

Iteration: [0/11] Loss 0.0013 Loss_CE 0.0013 Loss_var 0.0000 Pixel_Accu 0.9491 IoU 0.0000 2019-10-26 03:53 => Train Avg: Loss 0.0037 Loss_CE 0.0037 Loss_var 0.0000 Pixel_Accu 0.9559 IoU 0.0000 2019-10-26 03:54 => Val Avg: Loss 0.0000 Pixel_Acc 1.0000 IoU 0.0000 thank you for your help,i can finally train without errors,but do you know why the metrics except pixel_accu is always zero,this is a part my traning logs.And it the same problem with test.

[thinkInJava33]

sorry, Could you please provide the file or the values of mean and std

Sorry for the late reply. The mean_std.py is computed using the images in the training set. sorry, Could you please provide the file

this is the mean_std value calculated by myself，you can use it. normalize: [[0.78780321 0.5120167 0.78493782], [0.16766301 0.24838048 0.13225162]]

sorry to bother you again,do you know how to get the mutil organ dataset