I am new to image processing and I am applying keras_retinanet for object detection on my data set. My data set contains 20 classes, in each class, I have 1000 images. Each imag
