alexliyang

你好： 我想自己做一些数据集，能否详细描述下如何做出您那样的数据集呢？

is there any works to speed up the code ?

5

I try this on some color document image, and the output is not good. then I split the image to 3 channels and do three times, the output is good...

can you upload skipblocknet config about four different parameters?

1

In paper ,there are four different config about skipblocknet, ie. skipblocknet-xs,skipblocknet-s,skipblocknet-m and skipblocknet-l, they are just setting different width scale and mode?

do you try the code on color document image?

1

Thank you for your sharing these code. I try this work on color document image, it can not get the good output, do you have any idea?

how many epochs get the good trained model?

8

I use FiveK data set , resize the jpg to 480P , and use expertC as the target, but after 20 epochs , the test result is not good ....

I write some code about HDRNetCurves

1

I think use conv can replace google/hdrnet's ccm function block: # Color space change idtity = np.identity(nchans, dtype=np.float32) + np.random.randn(1).astype(np.float32)*1e-4 ccm = tf.get_variable('ccm', dtype=tf.float32, initializer=idtity) with tf.name_scope('ccm'): ccm_bias = tf.get_variable('ccm_bias',...

LPIENet pytorch code, when change x = torch.ones(1, 3, 16, 16) to x = torch.ones(1, 3, 2048, 2048) , then run it , get error Floating point exception (core dumped),...

After use this multi_gpu /cnn.py train a model ，how to save the pb model file ?

landmarks

1

can you support this model training with landmarks?