CenterCrop

class torchrl.envs.transforms.CenterCrop(w: int, h: int = None, in_keys: Sequence[NestedKey] | None = None, out_keys: Sequence[NestedKey] | None = None)

Crops the center of an image.

Parameters:

• w (int) – resulting width

• h (int, optional) – resulting height. If None, then w is used (square crop).

• in_keys (sequence of NestedKey, optional) – the entries to crop. If none is provided, ["pixels"] is assumed.

• out_keys (sequence of NestedKey, optional) – the cropped images keys. If none is provided, in_keys is assumed.

transform_observation_spec(observation_spec: TensorSpec) → TensorSpec

Transforms the observation spec such that the resulting spec matches transform mapping.

Parameters:

observation_spec (TensorSpec) – spec before the transform

Returns:

expected spec after the transform