我想从TFRecord读取数据

Question

我将图像日期保存到tfrecord中，但无法使用tensorflow数据集api对其进行解析。

我的环境

• Ubuntu 18.04
• Python 3.6.8
• Jupyter笔记本电脑
• Tensorflow 1.12.0

我通过以下代码保存了图像数据，

writer = tf.python_io.TFRecordWriter('training.tfrecord')

# X_train: paths to the image, y_train: labels (0 or 1)
for image_path, label in zip(X_train, y_train):
    image = cv2.imread(image_path)
    image = cv2.resize(image, (150, 150)) / 255.0
    ex = tf.train.Example(
        features = tf.train.Features(
            feature={
                'image' : tf.train.Feature(float_list = tf.train.FloatList(value=image.ravel())),
                'label' : tf.train.Feature(int64_list = tf.train.Int64List(value=[label]))
            }
        )
    )
    writer.write(ex.SerializeToString())
writer.close()

我试图从tfrecord文件中获取图像。

for record in tf.python_io.tf_record_iterator('test.tfrecord'):
    example = tf.train.Example()
    example.ParseFromString(record)

    img = example.features.feature['image'].float_list.value
    label = example.features.feature['label'].int64_list.value[0]

此方法有效。

但是当我使用Dataset API获取ML模型的图像时却没有。

def _parse_function(example_proto):
    features = {
        'label' : tf.FixedLenFeature((), tf.int64),
        'image' : tf.FixedLenFeature((), tf.float32)
    }
    parsed_features = tf.parse_single_example(example_proto, features)

    return parsed_features['image'], parsed_features['label']

def read_image(images, labels):
    label = tf.cast(labels, tf.int32)
    images = tf.cast(images, tf.float32)
    image = tf.reshape(images, [150, 150, 3])

# read the data
dataset = tf.data.TFRecordDataset('training.tfrecord')
dataset = dataset.map(_parse_function)
dataset = dataset.map(read_image) # <- ERROR!

错误提示是

ValueError: Cannot reshape a tensor with 1 elements to shape [150,150,3] (67500 elements) for 'Reshape' (op: 'Reshape') with input shapes: [], [3] and with input tensors computed as partial shapes: input[1] = [150,150,3].

我虽然导致此错误的原因是数组的形状有误，所以我确认了“数据集”的元素

<MapDataset shapes: ((), ()), types: (tf.float32, tf.int64)>

“数据集”变量没有数据。我不知道为什么会突然发作。

后记

结果，我尝试了Sharky的解决方案，

def parse(example_proto):
    features = {
        'label' : tf.FixedLenFeature((), tf.string, ''),
        'image' : tf.FixedLenFeature((), tf.string, '')
    }
    parsed_features = tf.parse_single_example(example_proto, features)
    img_shape = tf.stack([150, 150, 3])
    image = tf.decode_raw(parsed_features['image'], tf.float32)
    image = tf.reshape(image, img_shape)
    label = tf.decode_raw(parsed_features['label'], tf.int32)
    label = tf.reshape(label, tf.stack([1]))

    return image, label

我认为

有效。但是我无法从此MapDataset类型对象获取数组。该怎么做？

Answer 1

尝试使用单个解析函数

def parse(example_proto):
    features = {
        'label' : tf.FixedLenFeature((), tf.int64),
        'image' : tf.FixedLenFeature((), tf.string)
    }
    parsed_features = tf.parse_single_example(example_proto, features)
    img_shape = tf.stack([height, width, channel])
    image = tf.decode_raw(parsed_features['image'], tf.float32)
    image = tf.reshape(image, img_shape)
    label = tf.cast(parsed['label'], tf.int32)
    return image, label

好吧，似乎parse_single_example期望使用字符串类型而不是float。我建议像这样编码

def int64_feature(value):
    return tf.train.Feature(int64_list=tf.train.Int64List(value=[value]))


def bytes_feature(value):
    return tf.train.Feature(bytes_list=tf.train.BytesList(value=[value]))

writer = tf.python_io.TFRecordWriter('training.tfrecord')

for image_path, label in zip(X_train, y_train):
    image = cv2.imread(image_path)
    image = cv2.resize(image, (150, 150)) / 255.0
    img_raw = image.tostring()
    ex = tf.train.Example(features=tf.train.Features(feature={                                                                     
                        'image': bytes_feature(img_raw),
                        'label': int64_feature(label)
                         }))
    writer.write(ex.SerializeToString())
writer.close()