o oi@sXddlmZmZmZddlmZmZddlmZGdddej Z Gdddej Z d S) )ListTupleUnion)Tensornn)AugmentationSequentialcsveZdZdZdeddffdd ZdedefddZd ee e e e e fdee e e e e ffd d Z Z S) ManyToManyAugmentationDispatheraDispatches different augmentations to different inputs element-wisely. Args: augmentations: a list or a sequence of kornia AugmentationSequential modules. Examples: >>> import torch >>> input_1, input_2 = torch.randn(2, 3, 5, 6), torch.randn(2, 3, 5, 6) >>> mask_1, mask_2 = torch.ones(2, 3, 5, 6), torch.ones(2, 3, 5, 6) >>> aug_list = ManyToManyAugmentationDispather( ... AugmentationSequential( ... kornia.augmentation.ColorJiggle(0.1, 0.1, 0.1, 0.1, p=1.0), ... kornia.augmentation.RandomAffine(360, p=1.0), ... data_keys=["input", "mask",], ... ), ... AugmentationSequential( ... kornia.augmentation.ColorJiggle(0.1, 0.1, 0.1, 0.1, p=1.0), ... kornia.augmentation.RandomAffine(360, p=1.0), ... data_keys=["input", "mask",], ... ) ... ) >>> output = aug_list((input_1, mask_1), (input_2, mask_2)) augmentationsreturnNcst|j|||_dSN)super__init___check_consistencyr )selfr  __class__\/home/ubuntu/.local/lib/python3.10/site-packages/kornia/augmentation/container/dispatcher.pyr3s   z(ManyToManyAugmentationDispather.__init__cGs0t|D]\}}t|tstd|dqdS)N Please wrap your augmentations[`"`] with `AugmentationSequentials`.T) enumerate isinstancer ValueErrorrr iaugrrrr8s  z2ManyToManyAugmentationDispather._check_consistencyinputcGsddt||jDS)NcSsg|]\}}||qSrr).0inprrrr ?sz;ManyToManyAugmentationDispather.forward..)zipr rrrrrforward>sz'ManyToManyAugmentationDispather.forward)__name__ __module__ __qualname____doc__rrboolrrrrrr# __classcell__rrrrr s Br csxeZdZdZdddededdffdd Zdedefd d Zd ee e e fdee e e e e ffd dZ Z S)ManyToOneAugmentationDispatheraDispatches different augmentations to a single input and returns a list. Same `datakeys` must be applied across different augmentations. By default, with input (image, mask), the augmentations must not mess it as (mask, image) to avoid unexpected errors. This check can be cancelled with `strict=False` if needed. Args: augmentations: a list or a sequence of kornia AugmentationSequential modules. Examples: >>> import torch >>> input = torch.randn(2, 3, 5, 6) >>> mask = torch.ones(2, 3, 5, 6) >>> aug_list = ManyToOneAugmentationDispather( ... AugmentationSequential( ... kornia.augmentation.ColorJiggle(0.1, 0.1, 0.1, 0.1, p=1.0), ... kornia.augmentation.RandomAffine(360, p=1.0), ... data_keys=["input", "mask",], ... ), ... AugmentationSequential( ... kornia.augmentation.ColorJiggle(0.1, 0.1, 0.1, 0.1, p=1.0), ... kornia.augmentation.RandomAffine(360, p=1.0), ... data_keys=["input", "mask",], ... ) ... ) >>> output = aug_list(input, mask) T)strictr r+r Ncs$t||_|j|||_dSr )r rr+rr )rr+r rrrr`s   z'ManyToOneAugmentationDispather.__init__c Gst|D];\}}t|tstd|d|jr?|dkr?|j||djkr?td|dd|d|jd||djd qd S) NrrrrzDifferent `data_keys` between z and z elements, got .T)rrrrr+ data_keys RuntimeErrorrrrrrfs " z1ManyToOneAugmentationDispather._check_consistencyrcsfdd|jDS)Ncsg|]}|qSrr)rrrrrr rsz:ManyToOneAugmentationDispather.forward..)r r"rr/rr#qsz&ManyToOneAugmentationDispather.forward)r$r%r&r'rr(rrrrrrr#r)rrrrr*Bs  : r*N) typingrrrtorchrraugmentrModuler r*rrrrs  )