Tolstoy is an interactive art installation where the computer acts as a playful, imaginative partner in the authoring of stories. Visitors lie down in a cozy blanket fort and build a story together by speaking to the computer, adding new lines one at a time. Tolstoy reacts to each line and visualizes relevant subjects and objects as silhouettes, encouraging users steer the story in a new direction. It complements stories in a fun, unpredictable way.
Tolstoy is not limited to a few pre-programmed lines. It uses Web Speech API and NLTK to turn any input into text and extract meaning. Then, it scrapes Google Images for relevant visuals and turns these into starry outlines.
With: Subramanian Alagappan, Gayathri Gopalakrishnan
My role: Concepter, Researcher
Tech: NodeJS, Google Web Speech API, NLTK, three.js, ImageMagick, Web scraping