dltk.networks.autoencoder package

dltk.networks.autoencoder.convolutional_autoencoder module

dltk.networks.autoencoder.convolutional_autoencoder.convolutional_autoencoder_3d(inputs, num_convolutions=1, num_hidden_units=128, filters=(16, 32, 64), strides=((2, 2, 2), (2, 2, 2), (2, 2, 2)), mode='train', use_bias=False, activation=<function relu6>, kernel_initializer=<tensorflow.python.ops.init_ops.VarianceScaling object>, bias_initializer=<tensorflow.python.ops.init_ops.Zeros object>, kernel_regularizer=None, bias_regularizer=None)[source]
Convolutional autoencoder with num_convolutions on len(filters)
resolution scales. The downsampling of features is done via strided convolutions and upsampling via strided transpose convolutions. On each resolution scale s are num_convolutions with filter size = filters[s]. strides[s] determine the downsampling factor at each resolution scale.
Parameters:
  • inputs (tf.Tensor) – Input tensor to the network, required to be of rank 5.
  • num_convolutions (int, optional) – Number of convolutions per resolution scale.
  • num_hidden_units (int, optional) – Number of hidden units.
  • filters (tuple or list, optional) – Number of filters for all convolutions at each resolution scale.
  • strides (tuple or list, optional) – Stride of the first convolution on a resolution scale.
  • mode (str, optional) – One of the tf.estimator.ModeKeys strings: TRAIN, EVAL or PREDICT
  • use_bias (bool, optional) – Boolean, whether the layer uses a bias.
  • activation (optional) – A function to use as activation function.
  • kernel_initializer (TYPE, optional) – An initializer for the convolution kernel.
  • bias_initializer (TYPE, optional) – An initializer for the bias vector. If None, no bias will be applied.
  • kernel_regularizer (None, optional) – Optional regularizer for the convolution kernel.
  • bias_regularizer (None, optional) – Optional regularizer for the bias vector.
Returns:

dictionary of output tensors
Return type:

dict

Module contents