This application augments images to compensate for small image datasets.
User anotates the objects of interest and they automatically get randomly placed inside the same image with different color augmentations. All cloned objects are mirrored to increase variety and diversity in the dataset and robustness in the model.
Anotation files (YOLO anotations) are generated for every image enabling quick model training.