This repository contains the sRGBNoise model forked from Noise2NoiseFlow modified and pre-trained to generate sensor noise with a distribution similar to the noise of a Basler ace aCA640-750uc color camera.
The added sRGB_noise_modeling/add_noise.py script can augment noise-free RGB images with generated sensor noise using the sRGBNoise model. The model has been trained on images captured by a Basler ace aCA640-750uc color camera. One can use the script as follows:
python add_noise.py /input/image.png /output/noisy_image.png