BatchRandomTransform#

class torchaug.batch_transforms.BatchRandomTransform(p, inplace)[source]#

Abstract class to make a base class for all batch random transforms.

Parameters:

p (float) – Probability to apply the transform.
inplace (bool) – If True, perform inplace operation to save memory and time.

abstract apply_transform(imgs)[source]#

Function to perform transformation on the batch of images.

Note

Should be overridden by subclasses.

Parameters:: imgs (Tensor) – Batch of images to transform.
Return type:: Tensor
Returns:: The transformed batch of images.

forward(imgs)[source]#

Call apply_transform() on a randomly selected sub samples of passed images.

Parameters:: imgs (Tensor) – Images to randomly transform based on p.
Return type:: Tensor
Returns:: The randomly transformed images.