Pixel-level (Image-only) Transforms

Pixel-level transforms modify only the input image, leaving additional targets like masks, bounding boxes, and keypoints unchanged. Any target can be passed to these transforms; only the image will be modified.

Spatial-level (Dual) Transforms

Spatial-level transforms modify both the input image and associated targets (masks, bounding boxes, keypoints) simultaneously. Applying these transforms to unsupported targets will raise an error.

TransformImageMaskBBoxesKeypointsVolumeMask3D
Affine
AtLeastOneBBoxRandomCrop
BBoxSafeRandomCrop
CenterCrop
CoarseDropout
ConstrainedCoarseDropout
Crop
CropAndPad
CropNonEmptyMaskIfExists
D4
ElasticTransform
Erasing
FrequencyMasking
GridDistortion
GridDropout
GridElasticDeform
HorizontalFlip
Lambda
LongestMaxSize
MaskDropout
Morphological
NoOp
OpticalDistortion
OverlayElements
Pad
PadIfNeeded
Perspective
PiecewiseAffine
PixelDropout
RandomCrop
RandomCropFromBorders
RandomCropNearBBox
RandomGridShuffle
RandomResizedCrop
RandomRotate90
RandomScale
RandomSizedBBoxSafeCrop
RandomSizedCrop
Resize
Rotate
SafeRotate
ShiftScaleRotate
SmallestMaxSize
SquareSymmetry
ThinPlateSpline
TimeMasking
TimeReverse
Transpose
VerticalFlip
XYMasking

3D Transforms

3D transforms are designed for volumetric data (e.g., medical images) and operate on 3D volumes and their corresponding 3D masks. Applying these transforms to unsupported targets will raise an error.

TransformImageMaskBBoxesKeypointsVolumeMask3D
CenterCrop3D
CoarseDropout3D
CubicSymmetry
Pad3D
PadIfNeeded3D
RandomCrop3D