RandomChoice

class mmcv.transforms.RandomChoice(transforms: List[Dict | Callable[[Dict], Dict] | List[Dict | Callable[[Dict], Dict]]], prob: List[float] | None = None)

Process data with a randomly chosen transform from given candidates.

Parameters:

• transforms (list[list]) – A list of transform candidates, each is a sequence of transforms.

• prob (list[float], optional) – The probabilities associated with each pipeline. The length should be equal to the pipeline number and the sum should be 1. If not given, a uniform distribution will be assumed.

Examples

>>> # config
>>> pipeline = [
>>>     dict(type='RandomChoice',
>>>         transforms=[
>>>             [dict(type='RandomHorizontalFlip')],  # subpipeline 1
>>>             [dict(type='RandomRotate')],  # subpipeline 2
>>>         ]
>>>     )
>>> ]

transform(results: Dict) → Dict | None

Randomly choose a transform to apply.