Emotional Experience from Text
The project aim is to generate a piece of art given affective textual description.
The project aim is to generate a piece of art given affective textual description. Textual description may contain metaphors, similes or emotional statements. Our AI model accepts this text information and tries to reflect the mood visually through an artwork. We train our model using recently proposed ArtEmis dataset which contains over 420,000 utterances and 85,000 artworks from WikiArt dataset.
Artists: Kilichbek Haydarov & Mohamed Elhoseiny
Presented in NeurIPS Workshop on ML for Creativity and Design 2021.