This data set consists of a set of 870 images. Each image contains a hand making the shape of an ASL letter (with some variation). The purpose of this data set is to act as a sort of validation set to see how good the preprocessing and the model are, based on the following ASL Alphabet data set also available on Kaggle: https://www.kaggle.com/grassknoted/asl-alphabet. Content There are 30 images for each symbol, A-Z, delete, space, and nothing, making 870 images in total. The images are 200x200 8-bit photos to match the [asl-alphabet](https://www.kaggle.com/grassknoted/asl-alphabet) data set, and are organized in a folder structure to make it easy to use `flow_from_directory` in Keras. Acknowledgements The idea for this data set came about because of the [ASL Alphabet](https://www.kaggle.com/grassknoted/asl-alphabet) data set. Banner and thumbnail photo by [Sebastián León Prado on Unsplash](https://unsplash.com/photos/gbfRAzWR03E). Inspiration This data set is being used as part of a project for the W207 Applied Machine Learning class in UC Berkeley's master of information and data science program (MIDS).