Illumination

Lighting up words in sequence on a page creating poetry the machine reviles hidden texts within the text. This interactive installation accepts English articles the visitors bring & transports them to computer generated black-out poetry. The poems animated word by word & line by line fade away as new poetry is unveiled off the original article, leaving the freshly exposed interpretations to dissipate into the paper they're on.
The program equipped with statistical and grammatical rules & data was introduced to thousands of poetic works of authors ranging from Shakespeare to Yates, from Rabindranath Tagore to Lewis Carroll uses machine learning algorithms to build its own poetry in real time using the minimal vocabulary available in the unseen papers.

Illumination uses an Arduino microcontroller and hand made sensors to detect paper placing. A camera and a projector suspended in the lampshade above the table are used to take high resolution photos of the placed texts and light the paper. Tesseract OCR software generate text files for the images along with positioning information. Custom Python code generates the poetic texts which are then translated to projection mapping via Processing.py code animating the words out.

by Yonatan Ben-Simhon & Jack Kalish
Presented at GeekDown, at Ventana244, New York, 2011 and at Bloomfield Science Museum, Jerusalem, 2012