A MATLAB tutorial demonstrating how to use the NIMA (Neural Image Assessment) convolutional neural network to quantify image aesthetics and image quality — including downloading pretrained models, preprocessing and augmentation, training using the LIVE In the Wild dataset, modifying MobileNet-v2, and evaluating results with EMD, binary accuracy and correlation metrics.
6
6
6
8
5
7
9
9
7
6