hub.image_embedding_column

View source on GitHub

Uses a Module to get a dense 1-D representation from the pixels of images.

hub.image_embedding_column(
    key, module_spec, image_size=None
)

This feature column can be used on images, represented as float32 tensors of RGB pixel data in the range [0,1]. This can be read from a numeric_column() if the tf.Example input data happens to have decoded images, all with the same shape [height, width, 3]. More commonly, the input_fn will have code to explicitly decode images, resize them (possibly after performing data augmentation such as random crops etc.), and provide a batch of shape [batch_size, height, width, 3].

The result of this feature column is the result of passing its input through the module m instantiated from module_spec, as per result = m({"images": input}). The result must have dtype float32 and shape [batch_size, num_features] with a known value of num_features.

Example:

  image_column = hub.image_embedding_column("embeddings", "/tmp/image-module")
  feature_columns = [image_column, ...]
  estimator = tf.estimator.LinearClassifier(feature_columns, ...)
  height, width = hub.get_expected_image_size(image_column.module_spec)
  input_fn = ...  # Provides "embeddings" with shape [None, height, width, 3].
  estimator.train(input_fn, ...)