Dear Zhiyu,

could you please provide the pretrained model for SRCNN with RGB as input channel? The original paper only provides a model with Y as input.

Best regards
Andreas

I noticed that you missed the step of bluring the images between down-sampling and up-sampling. From prepare_train.py:

im_temp = scipy.misc.imresize(im_label, 1/scale, interp='bicubic')
im_input = scipy.misc.imresize(im_temp, size, interp='bicubic')

This still created distorted images, but I think it makes the task easier when the blur isn't applied.

The model's path is error. Modify ./checkpoint/ to ./checkpoint/srcnn_21/
You can just debug the code and find the problem.

Originally posted by @Nightree in #5 (comment)

Hi @Edwardlzy , I am replicating the same paper . I am having trouble pre-processing the input image.

This function is normalizing image converted to YCB_CR format .

Where in the code are you extracting the luminance value of YCB_CR format image?
Also, according to the paper, Gaussian blur needs to be added to the input image before rescaling it by factor of 3.

Could you provide the config file you used during training? Need this to convert ckpt to pb, thanks.

When I try to try the network with the padding = "SAME" in the code:
def conv2d(x, W):
return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding='VALID')
“””
return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding='SAME')
“””
in flie net.py,
and the it goes wrong:
in the code: l2_loss = tf.reduce_mean(tf.square(labels - model))
ValueError: Dimensions must be equal, but are 21 and 33 for 'sub' (op: 'Sub') with input shapes: [?,21,21,1], [?,33,33,1].
Can you tell me how to solve this problem? Thanks!

Hello! First of all, thank you for sharing your code on github. I had the honor to reproduce your program recently, but encountered a problem, as follows：

Traceback (most recent call last):
File "net.py", line 162, in
train_SRCNN(images)
File "net.py", line 92, in train_SRCNN
train_data, train_label = load_data(train_path)
File "net.py", line 24, in load_data
with h5py.File(path, 'r') as hf:
File "/root/anaconda3/lib/python3.7/site-packages/h5py/_hl/files.py", line 408, in init
swmr=swmr)
File "/root/anaconda3/lib/python3.7/site-packages/h5py/_hl/files.py", line 173, in make_fid
fid = h5f.open(name, flags, fapl=fapl)
File "h5py/_objects.pyx", line 54, in h5py._objects.with_phil.wrapper
File "h5py/_objects.pyx", line 55, in h5py._objects.with_phil.wrapper
File "h5py/h5f.pyx", line 88, in h5py.h5f.open
OSError: Unable to open file (unable to open file: name = './checkpoint/train.h5', errno = 2, error message = 'No such file or directory', flags = 0, o_flags = 0)

Can you help me solve this problem? Thank you for reading during your busy schedule and wish you a happy life!