-
An Application of Neural Embedding Models for Representing Artistic Periods
Rao Hamza Ali, Katie Rhodeghiero, Alexa Zuch, Saniya Syed, and Erik Linstead
We showcase visualizations created for art periods of Dalí, van Gogh, and Picasso by leveraging deep neural embedding models like word2vec to represent color features. First, the embedding vectors are generated for every color used in artworks of these painters. Next, t-distributed Stochastic Neighbor Embedding (t-SNE) is applied to generate a two-dimensional visualizations of the color space. Colors used in close proximity on the canvas are observed as compact clusters in the visualizations. These visualizations are termed as fingerprints, as they uniquely depict each art period of a painter, by highlighting the color palette used in their works. The authors further provide commentary on the artists’ art periods and how the fingerprints showcase their artistic evolution.
Below you may find selected books and book chapters from faculty in the Fowler School of Engineering, set to open in fall 2020.
Printing is not supported at the primary Gallery Thumbnail page. Please first navigate to a specific Image before printing.