Visual descriptions of artwork are an incredibly valuable tool that provide access to artworks for people with poor or no-vision. The WHO states that 2.2 billion people worldwide have some form of vision impairment. The descriptions allow everyone to understand what is happening in the work visually.
We have used the visual descriptions created by various institutions through the Coyote project to seed a machine learning model created by OpenAI called GPT-2.
To get started type the beginning of an artwork you would like to create into the text field below, then hit the begin exhibition link and enjoy your tour of a virtual gallery of your creation!
| begin exhibition ---> read another description ---> display your description ---> |