augmentation

There is a set of augmentations specific for medical imaging in the torchio.readthedocs.io.

The license there is Apache 2 => As I understand, we can use some of the transforms from there in Albumentations. Of course, we need to keep the link to the original implementation in the docstring.

For example:

• [RandomGhosting](https://torchio.readthedocs.io/transforms/augmentation.html#torc

This transform takes a fraction of the end or the start of the audio and treats that part as padding. We can implement several modes:

• constant (zero)
• edge - pads with the edge values of array
• wrap
• reflect

There are quite a few functions in the repo that need to be annotated.
This issue will be left open for external contributions. If you wish to contribute, please submit a PR for the same!

Thanks!
-J

The offset can be randomized, as long as the output has the specified length

Like https://github.com/Spijkervet/torchaudio-augmentations/blob/d044f9d020e12032ab9280acf5f34a337e72d212/torchaudio_augmentations/augmentations/random_resized_crop.py#L5

The idea is that one can have a chain of transforms, and some of them change the input length, but the final length should be fixed. That is where

Using the same concept used in keras segmentation example add pytorch segmentation example