weichen582 / RetinexNet

A Tensorflow implementation of RetinexNet
MIT License
678 stars 155 forks source link

Use of reduce_max in input #18

Open akshay-iyer opened 4 years ago

akshay-iyer commented 4 years ago

In the beginning of DeconNet function, there are these two ines:

_input_max = tf.reduce_max(input_im, axis=3, keepdims=True) input_im = concat([input_max, inputim])

Could you tell what it does and what is the use of the same?