This reposotory contains phases of the Computational Intelligence course at FUM. All the phases revolve around Digit-Five dataset.
Projects of this course are designed around the Digit-5 dataset, a multi-domain dataset consisting of MNIST, MNIST-M, SVHN, SYN, and USPS digit datasets. Images of each domain are digits from 0 to 9 with the following dimensions: 3 x 32 x 32. Each data point in the dataset includes an image of size 3 x 32 x 32, a corresponding digit label ranging from 0 to 9, and a domain label indicating which of the 5 domains the image belongs to (MNIST, MNIST-M, SVHN, SYN, or USPS). (Pictures of the MNIST dataset are resized to 32x32 and the single Black and white channel is copied three times to make the picture 3x32x32.)
A couple of samples from each domain are shown in the below figure. To explore and visualize more images from the dataset, run exploring_data/exploring_data.py.
