tf.gradients only returns [None]

Question

I tried to build my own deep dream algorithm with this code using the Inception Neural Network from Google:

import tensorflow as tf
import matplotlib.pyplot as plt
import numpy as np
#I am using the Kadenze CADL helper function ---> https://github.com/pkmital/CADL/tree/master/session-4/libs
import inception

img = np.random.rand(1,1920,1080,3)

net = inception.get_inception_model()
tf.import_graph_def(net['graph_def'], name='inception')
graph = tf.get_default_graph()
layer = graph.get_tensor_by_name('inception/mixed5b_pool_reduce_pre_relu:0')
gradient = tf.gradients(tf.reduce_mean(layer), img)
sess = tf.Session()
init = tf.global_variables_initializer()
iters = 1440

sess.run(init)
for i in range(iters):
    print(i+1)
    grad = sess.run(gradient[0])[0]
    img += grad
    plt.imshow(img[0])
    plt.savefig('output/'+str(i+1)+'.png')
    plt.close('all')

But the line tf.gradients(tf.reduce_mean(layer), img) only returns [None]. This (of course) causes an error. Can anyone tell me how to fix it?

`img` is a numpy array, and your gradients should be between 2 tensors with connected ops between them. — Vijay Mariappan, Aug 01 '17 at 20:17
I changed `np.random.rand(1,1920,1080,3)` to `tf.Variable(tf.random_normal([1,1920,1080,3]))` ,added the variable `softmax = sess.graph.get_tensor_by_name('inception/softmax2:0')` and made a connected operation between the gradient tensors with the line `sess.run(softmax, \ {'inception/input:0': sess.run(img)})`. But the error is still there. — Kay Jersch, Aug 02 '17 at 10:15

tf.gradients only returns [None]

0 Answers0