Datasets

1.1 Description

The Dot or No-Dot dataset is a collection of 1500 unique images, each measuring 32x32 pixels. This dataset is divided into two categories: images featuring a dot and those without one. The undotted images display a variety of randomly selected colored gradients, offering a visually diverse background. Conversely, the dotted images not only include different gradient backgrounds but also feature dots that vary in color, size, and placement, adding to the dataset's complexity. The motivation for the dataset can be found in the Personal Projects Section. Nevertheless, The dataset is accompanied by 1500 text files, one for each image. These files contain either a '0' or a '1'. A '0' signifies an image without a dot, while a '1' indicates the presence of a dot.

1.2 Images

1.3 Download

You can download the dataset here:

2.1 Description

Similar to the Dot or No-Dot dataset, the Tetris Dataset is a compilation of 1500 images, each either having a Tetris shape or none at all. Images with Tetris shapes can have one of five of the seven tetris shapes, namely all tetris shapes that are not mirror representations of each other. The txt files that come with the images contain the symbol representation of the tetris structure found in the image. For example, “S” would be one of the symbols. Nevertheless, the motivation for the highly varied tetris shape data set can be found in the Personal Projects Section.

2.2 Images

2.3 Download

You can download the dataset here: