d3rlpy.augmentation.image.VerticalFlip

class d3rlpy.augmentation.image.VerticalFlip(probability=0.1)

Vertical flip augmentation.

References

• Kostrikov et al., Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixels.

Parameters:probability (float) – probability to flip vertically.

probability

probability to flip vertically.

Type:float

Methods

get_params(deep=False)

Returns augmentation parameters.

Parameters:deep (bool) – flag to deeply copy objects. Returns:augmentation parameters. Return type:dict

get_type()

Returns augmentation type.

Returns:augmentation type. Return type:str

transform(x)

Returns vertically flipped image.

Parameters:x (torch.Tensor) – observation tensor. Returns:processed observation tensor. Return type:torch.Tensor

Attributes

TYPE = 'vertical_flip'