Image harmonization for deep learning model optimization

A system, comprising:

• a memory that stores computer executable components;
• and
• a processor that executes the computer executable components stored in the memory, wherein the computer executable components comprise:
• an image decomposition component that decomposes an input image into sub-images;
• and
• a harmonization component that harmonizes the sub-images with reference sub-images of respective reference images, resulting in transformation of the sub-images into modified sub-images, wherein the harmonization component harmonizes the sub-images with the reference sub-images using a weighting scheme for the reference sub-images determined based on degrees of similarity between the input image and the respective reference images.

A method, comprising:

• generating, by a system operatively coupled to a processor, ground-truth harmonized images for respective training images with different appearance variations, wherein the generating comprises employing at least one reference image and wherein the ground-truth harmonized images have a more similar appearance to the at least one reference image relative to the respective training images;
• training, by the system, a full-image harmonization model to transform the respective training images into the ground-truth harmonized images;
• and
• applying, by the system, the full-image harmonization model to a new input image to transform the new input image into a new harmonized image that has a more similar appearance to the at least one reference image relative to the new input image.

A method, comprising:

• generating, by a system operatively coupled to a processor, ground-truth sub-images for respective sub-images decomposed from training images, wherein the generating comprise employing at least one reference image;
• training, by the system, sub-image harmonization models to transform the sub-images into the ground-truth sub-images;
• applying, by the system, the sub-image machine learning models to transform new sub-images decomposed from an input image into new modified sub-images;
• and
• combining, by the system, the new modified sub-images to generate a harmonized image for the input image, the harmonized image having a more similar appearance to the at least one reference images relative to the input image.