Apply to speak at TensorFlow World. Deadline April 23rd. Propose talk

tf.contrib.gan.eval.preprocess_image

Aliases:

  • tf.contrib.gan.eval.classifier_metrics.preprocess_image
  • tf.contrib.gan.eval.preprocess_image
tf.contrib.gan.eval.preprocess_image(
    images,
    height=INCEPTION_DEFAULT_IMAGE_SIZE,
    width=INCEPTION_DEFAULT_IMAGE_SIZE,
    scope=None
)

Defined in tensorflow/contrib/gan/python/eval/python/classifier_metrics_impl.py.

Prepare a batch of images for evaluation.

This is the preprocessing portion of the graph from http://download.tensorflow.org/models/image/imagenet/inception-2015-12-05.tgz.

Note that it expects Tensors in [0, 255]. This function maps pixel values to [-1, 1] and resizes to match the InceptionV1 network.

Args:

  • images: 3-D or 4-D Tensor of images. Values are in [0, 255].
  • height: Integer. Height of resized output image.
  • width: Integer. Width of resized output image.
  • scope: Optional scope for name_scope.

Returns:

3-D or 4-D float Tensor of prepared image(s). Values are in [-1, 1].