opencv_transforms

This repository is intended as a faster drop-in replacement for Pytorch's Torchvision augmentations. This repo uses OpenCV for fast image augmentation for PyTorch computer vision pipelines. I wrote this code because the Pillow-based Torchvision transforms was starving my GPU due to slow image augmentation.

Requirements

• A working installation of OpenCV. Tested with OpenCV version 3.4.1

• Tested on Windows 10. There is evidence that OpenCV doesn't work well with multithreading on Linux / MacOS, for example num_workers >0 in a pytorch DataLoader. I haven't tried this on those systems.

Installation

• git clone https://github.com/jbohnslav/opencv_transforms.git

• Add to your python path

Usage

• from opencv_transforms import opencv_transforms as transforms

• From here, almost everything should work exactly as the original transforms.

Example: Image resizing

import numpy as np
image = np.random.randint(low=0, high=255, size=(1024, 2048, 3))
resize = transforms.Resize(size=(256,256))
image = resize(image)

Should be 1.5 to 10 times faster than PIL. See benchmarks

Performance

• Most transformations are between 1.5X and ~4X faster in OpenCV. Large image resizes are up to 10 times faster in OpenCV.

• To reproduce the following benchmarks, download the Cityscapes dataset.

• An example benchmarking file can be found in the notebook bencharming_v2.ipynb I wrapped the Cityscapes default directories with a HDF5 file for even faster reading.

The changes start to add up when you compose multiple transformations together. 

TODO

•  Initial commit with all currently implemented torchvision transforms

•  Cityscapes benchmarks

•  Make the resample flag on RandomRotation, RandomAffine actually do something

•  Speed up augmentation in saturation and hue. Currently, fastest way is to convert to a PIL image, perform same augmentation as Torchvision, then convert back to np.ndarray