Creating a data-set for a classification or segmentation task. If an annotation file is present, the annotations are also prepared. The data-set is created based on an image-set.

Helper Methods

setup_learner[source]

setup_learner(path)

Setup the learner with weights. path: The path to the folder containing the exported model file.

fix_odd_sides[source]

fix_odd_sides(img)

Fix odd image sizes. Odd image sizes can not be divided by 2. img: The image to fix as fastai image. return: the image reference.

class Inference[source]

Inference(path)

Base class for model inference.

class ClassificationInference[source]

ClassificationInference(path) :: Inference

Inference of a classification model. Supports single label and multi label classification.

class SegmentationInference[source]

SegmentationInference(path) :: Inference

Inference of a segmentation model.