张量流上的形状不兼容

Question

I am new in CNN,I am trying to make a CNN to classify image data set of handwritten English alphabet(az),(AZ) and numbers (0-9),which have 62 labels.Each image size is 30*30 pixels.I am following the steps as in a tutorial https://www.tensorflow.org/get_started/mnist/pros . When I run the model i get an error

tensorflow.python.framework.errors.InvalidArgumentError: Incompatible shapes: [40] vs. [10]

my batch size is 10,the error appears to be at the correct_prediction. The solution of the same problem found in Tensorflow Incompatable Shapes Error in Tutorial ,didn't fix my problem.Any help will be appreciated. The data set was compressed in pickle first,here are my codes.

import tensorflow as tf
import pandas as pd
import numpy as np
from sklearn.model_selection import train_test_split
X = []
y = []
import pickle

#load data from pickle
pickle_file = 'let.pickle'

with open(pickle_file, 'rb') as f:
    save = pickle.load(f)
    X = save['dataset']
    y = save['labels']

    del save  # hint to help gc free up memory

#normalise the features
X = (X - 255/2) / 255

# one hot encoding
y = pd.get_dummies(y)
y = y.values  # change to ndarray
y = np.float32(y)
X = np.float32(X)
Xtr, Xte, Ytr, Yte = train_test_split(X, y, train_size=0.7)

batch_size = 10

sess = tf.InteractiveSession()

x = tf.placeholder(tf.float32, shape=[None, 900])
y_ = tf.placeholder(tf.float32, shape=[None, 62])

def weight_variable(shape):
  initial = tf.truncated_normal(shape, stddev=0.1)
  return tf.Variable(initial)

def bias_variable(shape):
  initial = tf.constant(0.1, shape=shape)
  return tf.Variable(initial)

def conv2d(x, W):
  return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding='SAME')

def max_pool_2x2(x):
  return tf.nn.max_pool(x, ksize=[1, 2, 2, 1],
                        strides=[1, 2, 2, 1], padding='SAME')

W_conv1 = weight_variable([5, 5, 1, 32])
b_conv1 = bias_variable([32])

x_image = tf.reshape(x, [-1,30,30,1])

h_conv1 = tf.nn.relu(conv2d(x_image, W_conv1) + b_conv1)
h_pool1 = max_pool_2x2(h_conv1)

W_conv2 = weight_variable([5, 5, 32, 64])
b_conv2 = bias_variable([64])

h_conv2 = tf.nn.relu(conv2d(h_pool1, W_conv2) + b_conv2)
h_pool2 = max_pool_2x2(h_conv2)

W_fc1 = weight_variable([4 * 4 * 64, 1024])
b_fc1 = bias_variable([1024])

h_pool2_flat = tf.reshape(h_pool2, [-1, 4*4*64])
h_fc1 = tf.nn.relu(tf.matmul(h_pool2_flat, W_fc1) + b_fc1)

keep_prob = tf.placeholder(tf.float32)
h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob)

W_fc2 = weight_variable([1024, 62])
b_fc2 = bias_variable([62])

y_conv = tf.matmul(h_fc1_drop, W_fc2) + b_fc2

cross_entropy = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(labels=y_, logits=y_conv))
train_step = tf.train.AdamOptimizer(1e-4).minimize(cross_entropy)
correct_prediction = tf.equal(tf.argmax(y_conv,1), tf.argmax(y_,1))

accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))
sess.run(tf.global_variables_initializer())
for i in range(20000):
    offset = (i * batch_size) % (Ytr.shape[0] - batch_size)
    batch_x = Xtr[offset:(offset + batch_size), :]
    batch_y = Ytr[offset:(offset + batch_size), :]
    if i%100 == 0:
        train_accuracy = accuracy.eval(feed_dict={x:batch_x , y_: batch_y, keep_prob: 1.0})
        print("step %d, training accuracy %g"%(i, train_accuracy))
    train_step.run(feed_dict={x: Xtr[offset:(offset + batch_size), :], y_: Ytr[offset:(offset + batch_size), :], keep_prob: 0.5})

print("test accuracy %g"%accuracy.eval(feed_dict={x: Xte, y_: Yte, keep_prob: 1.0}))

Answer 1

Take a look at these lines of code:

cross_entropy = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(labels=y_, logits=y_conv))
correct_prediction = tf.equal(tf.argmax(y_conv,1), tf.argmax(y_,1))

If labels is a list with the index of the correct prediction (for example: [1,32,13, ...] ) the softmax_cross_entropy_with_logits function is correct. This means that the error is in these lines. I put in a comment that they do:

tf.argmax(y_conv,1) # Takes the max index of logits
tf.argmax(y_,1) # Takes the max index of ???. 

Although I did not test it, replacing it with this line should work:

correct_prediction = tf.equal(tf.argmax(y_conv,1), y_)

Let me know when you fixed it :D

Answer 2

I change the dimensions of the input to the fully connected layer from 4 to 8,its working now

W_fc1 = weight_variable([8 * 8 * 64, 1024])
b_fc1 = bias_variable([1024])

#the input should be shaped/flattened
h_pool2_flat = tf.reshape(h_pool2, [-1, 8*8*64])
h_fc1 = tf.nn.relu(tf.matmul(h_pool2_flat, W_fc1) + b_fc1)