Python Tensorflow和cifar 10，测试单个图像

我试图用tensorflow的cifar-10预测单个图像的类别

我找到了此代码，但失败，出现以下错误：

Assign需要两个张量的形状匹配。左侧形状=[18384]右侧形状=[2304384] 我知道这是因为该批次的规模只有1。（使用expand_dims，我创建了一个假批次。）

但我不知道怎么解决这个问题

我到处找，但没有找到解决办法。。 提前谢谢

from PIL import Image
import tensorflow as tf
from tensorflow.models.image.cifar10 import cifar10
width = 24
height = 24

categories =  ["airplane","automobile","bird","cat","deer","dog","frog","horse","ship","truck" ]

filename = "path/to/jpg" # absolute path to input image
im = Image.open(filename)
im.save(filename, format='JPEG', subsampling=0, quality=100)
input_img = tf.image.decode_jpeg(tf.read_file(filename), channels=3)
tf_cast = tf.cast(input_img, tf.float32)
float_image = tf.image.resize_image_with_crop_or_pad(tf_cast, height, width)
images = tf.expand_dims(float_image, 0)
logits = cifar10.inference(images)
_, top_k_pred = tf.nn.top_k(logits, k=5)
init_op = tf.initialize_all_variables()
with tf.Session() as sess:
saver = tf.train.Saver()
ckpt = tf.train.get_checkpoint_state('/tmp/cifar10_train')
if ckpt and ckpt.model_checkpoint_path:
    print("ckpt.model_checkpoint_path ", ckpt.model_checkpoint_path)
    saver.restore(sess, ckpt.model_checkpoint_path)
else:
    print('No checkpoint file found')
    exit(0)
sess.run(init_op)
_, top_indices = sess.run([_, top_k_pred])
for key, value in enumerate(top_indices[0]):
    print (categories[value] + ", " + str(_[0][key]))

编辑

我尝试放置占位符，第一个形状中没有占位符，但出现以下错误： 新变量（local3/weights）的形状必须完全定义，但实际上是（？，384）

现在我真的迷路了。。 以下是新代码：

from PIL import Image
import tensorflow as tf
from tensorflow.models.image.cifar10 import cifar10
import itertools
width = 24
height = 24

categories = [ "airplane","automobile","bird","cat","deer","dog","frog","horse","ship","truck" ]

filename = "toto.jpg" # absolute path to input image
im = Image.open(filename)
im.save(filename, format='JPEG', subsampling=0, quality=100)
x = tf.placeholder(tf.float32, [None, 24, 24, 3])
init_op = tf.initialize_all_variables()
with tf.Session() as sess:
 # Restore variables from training checkpoint.
    input_img = tf.image.decode_jpeg(tf.read_file(filename), channels=3)
    tf_cast = tf.cast(input_img, tf.float32)
    float_image = tf.image.resize_image_with_crop_or_pad(tf_cast, height, width)
    images = tf.expand_dims(float_image, 0)
    i = images.eval()
    print (i)
    sess.run(init_op, feed_dict={x: i})
    logits = cifar10.inference(x)
    _, top_k_pred = tf.nn.top_k(logits, k=5)
    variable_averages = tf.train.ExponentialMovingAverage(
        cifar10.MOVING_AVERAGE_DECAY)
    variables_to_restore = variable_averages.variables_to_restore()
    saver = tf.train.Saver(variables_to_restore)
    ckpt = tf.train.get_checkpoint_state('/tmp/cifar10_train')
    if ckpt and ckpt.model_checkpoint_path:
        print("ckpt.model_checkpoint_path ", ckpt.model_checkpoint_path)
        saver.restore(sess, ckpt.model_checkpoint_path)
    else:
        print('No checkpoint file found')
        exit(0)
    _, top_indices = sess.run([_, top_k_pred])
    for key, value in enumerate(top_indices[0]):
        print (categories[value] + ", " + str(_[0][key]))

---

我认为这是因为通过

tf.variable

或

tf.get\u variable

获取的变量必须具有完整定义的形状。您可以检查代码并给出完整定义的形状