Home BlogDataset Training Multimodal AI Systems

Training Multimodal AI Systems

by Morgan Stevens
by

An independent team of researchers has created a dataset of wordplay puzzles that require users to add or subtract letters from words to identify a phrase. It contains 333 puzzles from 13 categories, such as major cities and food. Researchers can use the dataset to improve multimodal AI systems that can decipher clued phrases found in text and images. 

Get the data.

Image credit: Flickr user Rob White

You may also like

Show Buttons
Hide Buttons