NonParametricRegressorTrainer

class menpo.fit.regression.trainer.NonParametricRegressorTrainer(reference_shape, regression_type=<function mlr at 0x7f5c2da62cf8>, regression_features=<function sparse_hog at 0x7f5c2da626e0>, patch_shape=(16, 16), noise_std=0.04, rotation=False, n_perturbations=10)

Bases: RegressorTrainer

Class for training a Non-Parametric Regressor.

Parameters:

reference_shape : PointCloud

The reference shape that will be used.

regression_type : function, optional

A function that defines the regression technique to be used. Examples of such closures can be found in Functions

regression_features : None or string or function, optional

The features that are used during the regression.

If None, no feature representation will be computed from the original image.

If string or function, the feature representation will be computed in the following way:

If string, the feature representation will be extracted by executing:

feature_image = getattr(image.features, feature_type)()

For this to work properly feature_type needs to be one of Menpo’s standard image feature methods. Note that, in this case, the feature computation will be carried out using its default options.

Non-default feature options and new experimental feature can be used by defining a closure. In this case, the closure must define a function that receives as an input an image and returns a particular feature representation of that image. For example:

def igo_double_from_std_normalized_intensities(image)
    image = deepcopy(image)
    image.normalize_std_inplace()
    return image.feature_type.igo(double_angles=True)

See ImageFeatures for details more details on Menpo’s standard image features and feature options. See Features Functions for non standard features definitions.

patch_shape : tuple, optional

The shape of the patches that will be extracted.

noise_std : float, optional

The standard deviation of the gaussian noise used to produce the training shapes.

rotation : boolean, optional

Specifies whether ground truth in-plane rotation is to be used to produce the training shapes.

n_perturbations : int, optional

Defines the number of perturbations that will be applied to the training shapes.

delta_ps(gt_shape, perturbed_shape)

Method to generate the delta_ps for the regression.

Parameters:

gt_shape : PointCloud

The ground truth shape.

perturbed_shape : PointCloud

The perturbed shape.

features(image, shape)

Method that extracts the features for the regression, which in this case are patch based.

Parameters:

image : MaskedImage

The current image.

shape : PointCloud

The current shape.

perturb_shapes(gt_shape)

Perturbs the given shapes. The number of perturbations is defined by n_perturbations.

Parameters:

gt_shape : PointCloud list

List of the shapes that correspond to the images. will be perturbed.

Returns:

perturbed_shapes : PointCloud list

List of the perturbed shapes.

train(images, shapes, perturbed_shapes=None, verbose=False, **kwargs)

Trains a Regressor given a list of landmarked images.

Parameters:

images : list of MaskedImage

The set of landmarked images from which to train the regressor.

shapes : PointCloud list

List of the shapes that correspond to the images.

perturbed_shapes : PointCloud list, optional

List of the perturbed shapes used for the regressor training.

verbose : boolean, optional

Flag that controls information and progress printing.

Returns:

regressor : :map:`Regressor`

A regressor object.

Raises:

ValueError :

The number of shapes must be equal to the number of images.

ValueError :

The number of perturbed shapes must be equal or multiple to the number of images.

algorithm

Returns the algorithm name.