小编aur*_*oua的帖子

我使用tensorflow对象检测API中的代码生成了pascal voc 2007 tfrecords文件。我使用tf.contrib.data.DatasetAPI从tfrecords中读取数据。我在没有tf.contrib.data.DatasetAPI的情况下尝试了mehtod ，并且代码可以正常运行，但是更改为tf.contrib.data.DatasetAPI后无法正常工作。

没有tf.contrib.data.Dataset以下代码：

import tensorflow as tf

if __name__ == '__main__':
    slim_example_decoder = tf.contrib.slim.tfexample_decoder

    features = {"image/height": tf.FixedLenFeature((), tf.int64, default_value=1),
                "image/width": tf.FixedLenFeature((), tf.int64, default_value=1),
                "image/filename": tf.FixedLenFeature((), tf.string, default_value=""),
                "image/source_id": tf.FixedLenFeature((), tf.string, default_value=""),
                "image/key/sha256": tf.FixedLenFeature((), tf.string, default_value=""),
                "image/encoded": tf.FixedLenFeature((), tf.string, default_value=""),
                "image/format": tf.FixedLenFeature((), tf.string, default_value="jpeg"),
                "image/object/bbox/xmin": tf.VarLenFeature(tf.float32),
                "image/object/bbox/xmax": tf.VarLenFeature(tf.float32),
                "image/object/bbox/ymin": tf.VarLenFeature(tf.float32),
                "image/object/bbox/ymax": tf.VarLenFeature(tf.float32),
                "image/object/class/text": tf.VarLenFeature(tf.string),
                "image/object/class/label": tf.VarLenFeature(tf.int64),
                "image/object/difficult": tf.VarLenFeature(tf.int64),
                "image/object/truncated": tf.VarLenFeature(tf.int64),
                "image/object/view": tf.VarLenFeature(tf.int64)}
    items_to_handlers = {
        'image': slim_example_decoder.Image(
            image_key='image/encoded', …

我正在使用以下代码生成 tfrecords 文件。

  def generate_tfrecords(data_path, labels, name):
      """Converts a dataset to tfrecords."""
      filename = os.path.join(args.tfrecords_path, name + '.tfrecords')
      writer = tf.python_io.TFRecordWriter(filename)
for index, data in enumerate(data_path):
      with tf.gfile.GFile(data, 'rb') as fid:
        encoded_jpg = fid.read()
        print(len(encoded_jpg))   # 17904
      encoded_jpg_io = io.BytesIO(encoded_jpg)
      image = pil.open(encoded_jpg_io)
      image = np.asarray(image)
      print(image.shape)    # 112*112*3
      example = tf.train.Example(features=tf.train.Features(feature={
        'height': _int64_feature(int(image.shape[0])),
        'width': _int64_feature(int(image.shape[1])),
        'depth': _int64_feature(int(3)),
        'label': _int64_feature(int(labels[index])),
        'image_raw': _bytes_feature(encoded_jpg)}))
      writer.write(example.SerializeToString())
    writer.close()

在上面的代码中，encoded_jpg有长度17904，图像有112*112*3不一致的形状。

当我使用以下代码解析 tfrecords 时：

def _parse_function(example_proto):
     features = {'height':  tf.FixedLenFeature((), …