8.1.1.3.1.24. blueoil.datasets.ytfaces

8.1.1.3.1.24.1. Module Contents

8.1.1.3.1.24.1.1. Classes

YoutubeFacialLandmarks

Youtube Facial landmarks detection dataset. This dataset is taken from:

class blueoil.datasets.ytfaces.YoutubeFacialLandmarks(subset='train', batch_size=10, *args, **kwargs)

Bases: blueoil.datasets.base.KeypointDetectionBase

Youtube Facial landmarks detection dataset. This dataset is taken from: https://github.com/jerrychen44/face_landmark_detection_pytorch

classes = ['face']
num_classes
available_subsets = ['train', 'validation']
extend_dir = ytfaces
_load_csv(self)

Read items from JSON files

__getitem__(self, item)

Get an item given index.

Parameters

item – int, index.

Returns

a numpy array of shape (height, width, 3). joints: a numpy array of shape (68, 3), which has coordinates in image.

Return type

image

__len__(self)

returns the number of items in the dataset.

property num_per_epoch(self)

Returns the number of datas in the data subset.