Transforms

New transform Mosaic

Generalization of Mosaic from Ultralitics and YOLO4, and works per image an not on "batch" => can choose what additional images to pass, could be hard or rare classes.

by @Shysto and @ternaus

Core

SomeOf

Changed functionality to a more intuituve

Now it works as:

1. Select n transforms with equal probability
2. Apply each of them with the probability of each transform

Passing bounding bbox labels

Removed to pass labels when apply to bounding boxes.

In [9]: bboxes = np.array([[0.2, 0.2, 0.4, 0.4], [0.3, 0.4, 0.7, 0.9]])

In [10]: transform = A.Compose([A.HorizontalFlip(p=1)], bbox_params={"format": "albumentations"})

In [11]: image = np.random.rand(640, 640, 3)

In [12]: transformed = transform(image=image, bboxes=bboxes)

=> we can just pass coordinates, without bounding box labels

Speedups

When applied to uint images on 1 CPU core Albumentations outperforms Kornia and torchvision: Image benchmark

But when we compare:

• Videos
• Albumentations on 1 CPU core vs kornia and torchvision on GTX 4090

Albumentations has a lot to improve. Benchmark on videos

=> Speedups on videos in this release:

• HorizontalFlip
• VerticalFlip
• RandomRotate90
• ChannelShuffle
• Crop
• RandomCrop
• CenterCrop

Bugfixes

• Bugfix in RandomRain, drop_length was not used before
• BugFix in ElasticTransform Added back fill and fill_mask parameters as for extreme deformations they become useful. Also switched kernels for exact and approximate mode
• Fix in docsting by @nicolasj92
• Cleanup in composition module by @dmsy4