__init__

UNet.__init__(input_features, quantiles, n_features=32, n_levels=4, skip_connection=None)[source]
Parameters:
  • input_features (int) – The number of channels of the input image.

  • quantiles (np.array) – Array containing the quantiles to predict.

  • n_features – The number of channels of the first convolution block.

  • n_level – The number of down-sampling steps.

  • skip_connection – Whether or not to include skip connections in each block.