fid-cnn's Introduction

idcnn - tensorflow

SAR Image Despeckling Using a Convolutional Neural Network by Tensorflow

Usage -- Change the parameters in code to make it.

The parameters for train

MODE = "train"
OUTDIR = "output_train"
MAX_EPOCH = 200
INPUTDIR = "data_train"

The parameters for test

MODE = "test"
OUTDIR = "output_test"
INPUTDIR = "data_test"
CHECKPOINT = "output_train"

NOTICE:

The dataset need to be grey channel rather than rgb.

fid-cnn's People

Contributors

Stargazers

Watchers

fid-cnn's Issues

制作的数据和作者的不符？

左边是作者论文给出的一张例图，右边是用你的方法（代码如下MATLAB）加噪后的图，直方图明显不一致，请问是什么问题？

代码：

noise_arr = (gamrnd(25, 0.04, h, w) - 1);
im_n = uint8(im2double(im) .* (1 + noise_arr) * 255);

图像整体变暗

尊敬的作者你好，请问下怎么运行这个，我设置好了INPUT，OUTPUT路径还是没有训练起来，报错了，报错内容ValueError:Tried to convert 'values' to a tensor and failed. Error: None values not supported.

VErsion

Please indicate the packages used and also the versions
Thanks

Nan in summary histogram

Thanks for the code. I tried to execute it got the following error. can you please tell me why this problem occurred.

WARNING:tensorflow:From :10: Supervisor.init (from tensorflow.python.training.supervisor) is deprecated and will be removed in a future version.
Instructions for updating:
Please switch to tf.train.MonitoredTrainingSession
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
INFO:tensorflow:Starting standard services.
INFO:tensorflow:Starting queue runners.
parameter_count = 223104
progress epoch 1 step 50 image/sec 0.6 remaining 5695m
loss [[[[nan]
[nan]
[nan]
.....
INFO:tensorflow:Error reported to Coordinator: <class 'tensorflow.python.framework.errors_impl.InvalidArgumentError'>, Nan in summary histogram for: decode_3/batchnorm/offset/values
[[node decode_3/batchnorm/offset/values (defined at :12) ]]

Recommend Projects