Skip to content

fast image augmentation library and easy to use wrapper around other libraries

License

Notifications You must be signed in to change notification settings

linghai06/albumentations

Repository files navigation

Albumentations

Build Status Documentation Status

  • Great fast augmentations based on highly-optimized OpenCV library
  • Super simple yet powerful interface for different tasks like (segmentation, detection, etc.)
  • Easy to customize
  • Easy to add other frameworks

Vladimir_Iglovikov

Authors

Alexander Buslaev

Alex Parinov

Vladimir Iglovikov

Example usage

from albumentations import (
    CLAHE, RandomRotate90, Transpose, ShiftScaleRotate, Blur, OpticalDistortion, 
    GridDistortion, HueSaturationValue, IAAAdditiveGaussianNoise, GaussNoise, MotionBlur, 
    MedianBlur, IAAPiecewiseAffine, IAASharpen, IAAEmboss, RandomContrast, RandomBrightness, 
    Flip, OneOf, Compose
)
import numpy as np

def strong_aug(p=0.5):
    return Compose([
        RandomRotate90(),
        Flip(),
        Transpose(),
        OneOf([
            IAAAdditiveGaussianNoise(),
            GaussNoise(),
        ], p=0.2),
        OneOf([
            MotionBlur(p=0.2),
            MedianBlur(blur_limit=3, p=0.1),
            Blur(blur_limit=3, p=0.1),
        ], p=0.2),
        ShiftScaleRotate(shift_limit=0.0625, scale_limit=0.2, rotate_limit=45, p=0.2),
        OneOf([
            OpticalDistortion(p=0.3),
            GridDistortion(p=0.1),
            IAAPiecewiseAffine(p=0.3),
        ], p=0.2),
        OneOf([
            CLAHE(clip_limit=2),
            IAASharpen(),
            IAAEmboss(),
            RandomContrast(),
            RandomBrightness(),
        ], p=0.3),
        HueSaturationValue(p=0.3),
    ], p=p)

image = np.ones((300, 300, 3), dtype=np.uint8)
mask = np.ones((300, 300), dtype=np.uint8)
whatever_data = "my name"
augmentation = strong_aug(p=0.9)
data = {"image": image, "mask": mask, "whatever_data": whatever_data, "additional": "hello"}
augmented = augmentation(**data)
image, mask, whatever_data, additional = augmented["image"], augmented["mask"], augmented["whatever_data"], augmented["additional"]

See example.ipynb

Installation

You can use pip to install albumentations:

pip install albumentations

If you want to get the latest version of the code before it is released on PyPI you can install the library from GitHub:

pip install -U git+https://github.com/albu/albumentations

Documentation

The full documentation is available at albumentations.readthedocs.io.

Demo

You can use this Google Colaboratory notebook to adjust image augmentation parameters and see the resulting images.

Migrating from torchvision to albumentations

Migrating from torchvision to albumentations is simple - you just need to change a few lines of code. Albumentations has equivalents for common torchvision transforms as well as plenty of transforms that are not presented in torchvision. migrating_from_torchvision_to_albumentations.ipynb shows how one can migrate code from torchvision to albumentations.

Working with non-8-bit images

example_16_bit_tiff.ipynb shows how albumentations can be used to work with non-8-bit images (such as 16-bit and 32-bit TIFF images).

Object detection tasks and bounding boxes

example_bboxes.ipynb shows how albumentations can be used to work with bounding boxes.

Benchmarking results

To run the benchmark yourself follow the instructions in benchmark/README.md

Results for running the benchmark on first 2000 images from the ImageNet validation set using an Intel Core i7-7800X CPU. All times are in seconds, lower is better.

albumentations imgaug torchvision
(Pillow backend)
torchvision
(Pillow-SIMD backend)
Keras
RandomCrop64 0.0017 - 0.0182 0.0182 -
PadToSize512 0.2413 - 2.493 2.3682 -
HorizontalFlip 0.7765 2.2299 0.3031 0.3054 2.0508
VerticalFlip 0.178 0.3899 0.2326 0.2308 0.1799
Rotate 3.8538 4.0581 16.16 9.5011 50.8632
ShiftScaleRotate 2.0605 2.4478 18.5401 10.6062 47.0568
Brightness 2.1018 2.3607 4.6854 3.4814 9.9237
ShiftHSV 10.3925 14.2255 34.7778 27.0215 -
ShiftRGB 2.6159 2.1989 - - 3.0598
Gamma 1.4832 - 1.1397 1.1447 -
Grayscale 1.2048 5.3895 1.6826 1.2721 -

Contributing

  1. Clone the repository:
git clone [email protected]:albu/albumentations.git
cd albumentations
  1. Install the library in development mode:
pip install -e .[tests]
  1. Run tests:
pytest

Building the documentation

  1. Go to docs/ directory
cd docs
  1. Install required libraries
pip install -r requirements.txt
  1. Build html files
make html
  1. Open _build/html/index.html in browser.

Alternatively, you can start a web server that rebuilds the documentation automatically when a change is detected by running make livehtml

Comments

In some systems, in the multiple GPU regime PyTorch may deadlock the DataLoader if OpenCV was compiled with OpenCL optimizations. Adding the following two lines before the library import may help. For more details pytorch/pytorch#1355

cv2.setNumThreads(0)	
cv2.ocl.setUseOpenCL(False)

Thanks:

Special thanks to @creafz for refactoring, documentation, tests, CI and benchmarks. Awesome work!

About

fast image augmentation library and easy to use wrapper around other libraries

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Python 99.7%
  • Shell 0.3%