CODEWITHSUNDEEP
pythontensorflowmachine-learningpredict
buydadip
buydadip

Reputation: 9407

CNN model not able to make prediction

I trained a CNN model successfully, however I am getting errors when I feed images to the model for it to predict the labels.

This is my model (I am restoring it with saver.restore)...

# load dataset
mnist = input_data.read_data_sets("/tmp/data/", one_hot=True)

# interactive session
sess = tf.InteractiveSession()

# data and labels placeholder
x = tf.placeholder(tf.float32, shape=[None, 784])
y = tf.placeholder(tf.float32, shape=[None, 10])

# 32 filters of size 5x5 and 32 biases,
# the filters are used to create 32 feature maps
W_conv1 = weight_variable([5, 5, 1, 32])
b_conv1 = bias_variable([32])

x_img = tf.reshape(x, [-1, 28, 28, 1])

# first layer activated using a Relu activation function
conv1 = tf.nn.relu(conv2d(x_img, W_conv1) + b_conv1)
pool1 = max_pool_2x2(conv1)

# 64 filters of size 5x5
W_conv2 = weight_variable([5, 5, 32, 64])
b_conv2 = bias_variable([64])

# second layer
conv2 = tf.nn.relu(conv2d(pool1, W_conv2) + b_conv2)
pool2 = max_pool_2x2(conv2)

# fully connected layer with 1024 neurons
W_fully = weight_variable([7 * 7 * 64, 1024])
b_fully = bias_variable([1024])

pool2flat = tf.reshape(pool2, [-1, 7 * 7 * 64])
fully = tf.nn.relu(tf.matmul(pool2flat, W_fully) + b_fully)

# dropout layer removes dead neurons
prob_drop = tf.placeholder(tf.float32)
dropout = tf.nn.dropout(fully, prob_drop)

# readout layer that will return the raw values
# of our predictions
W_readout = weight_variable([1024, 10])
b_readout = bias_variable([10])

y_conv = tf.matmul(dropout, W_readout) + b_readout

# loss function
cross_entropy = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=y_conv, labels=y))

# restore the trained CNN model
saver = tf.train.Saver()
saver.restore(sess, "/tmp/model2.ckpt")

y_conv is the predictor.

The model is trained on the mnist dataset, now I have an image of a number and I want the model to tell me what it thinks it is in terms of accuracy. I tried the following...

prediction = tf.argmax(y_conv, 1)
print(sess.run(prediction, feed_dict={x:two_images[0]}))

After feeding the image two_images[0] to the model, I got the following error...

ValueError: Cannot feed value of shape (784,) for Tensor 'Placeholder:0', which has shape '(?, 784)'

So I fixed it by doing the following...

prediction = tf.argmax(y_conv, 1)
print(sess.run(prediction, feed_dict={x:two_images[0].reshape((1, 784))}))

But now I am getting a whole bunch of errors that I cannot decipher...

InvalidArgumentError (see above for traceback): You must feed a value for placeholder tensor 'Placeholder_2' with dtype float [[Node: Placeholder_2 = Placeholderdtype=DT_FLOAT, shape=, _device="/job:localhost/replica:0/task:0/device:CPU:0"]]

I am not sure what I am doing wrong.

EDIT

This is how I populate the variable two_images...

# extract the indices of the number 2
two_idxs_list = np.where(mnist.test.labels[:, 2].astype(int) == 1)
two_idxs = two_idxs_list[0][:10]

# use the indices to extract the images of 2 and their corresponding label
two_images = mnist.test.images[two_idxs]
two_labels = mnist.test.labels[two_idxs]

Upvotes: 0

Views: 225

Answers (1)

Peter Szoldan
Peter Szoldan

Reputation: 4868

Okay with the code added I was able to test in on my machine. The issue is that your network expects two inputs, one image, and a label. Even if you only do inference, you have to supply an input, maybe just some zeroes? Obviously the loss calculation will be wrong, but you're not interested in that, only in the prediction. So your sess.run line should be:

print( sess.run( prediction, feed_dict= {
    x: two_images[0].reshape((1, 784)),
    y: np.zeros( shape = ( 1, 10 ), dtype = np.float32 ) } ) )

Upvotes: 1

Related Questions