generators¶

Generators to continusously resolve features.

Classes¶

class deeptrack.generators.Generator¶

Base class for a generator.

Generators continously update and resolve features, and allow other frameworks to continuously access new features.

generate(feature, label_function=None, batch_size=1, shuffle_batch=True, ndim=4)¶

Create generator instance.

Parameters

feature (Feature) – The feature to resolve images from.
label_function (Callable[Image] -> array_like) – Function that returns the label corresponding to an image.
batch_size (int) – Number of images per batch.
shuffle_batch (bool) – If True, the batches are shuffled before outputting.
ndim (int) – Expected number of dimensions of the output.