DataLoader

class maite.protocols.multiobject_tracking.DataLoader(*args, **kwargs)

A dataloader protocol for the multi-object tracking AI problem providing batch-level data access.

Implementers must provide an iterable object (returning an iterator via the __iter__ method) that yields tuples containing batches of data. These tuples contain types Sequence [VideoStream], Sequence [MultiobjectTrackingTarget], and Sequence [DatumMetadata], which correspond to model input batch, model target type batch, and a datum metadata batch.

Note: Unlike Dataset, this protocol does not require indexing support, only iterating.

Methods

__iter__ -> Iterator[tuple[Sequence[VideoStream], Sequence[MultiobjectTrackingTarget], Sequence[DatumMetadata]]]

Return an iterator over batches of data, where each batch contains a tuple of of model input batch (as Sequence [VideoStream]), model target batch (as Sequence [MultiobjectTrackingTarget]), and batched datum-level metadata (as Sequence [DatumMetadata]), respectively.