Artist Ross Goodwin uses a Neural Network trained on the collected publications of the Teachers College Record to create a version of his word.camera image captioning piece specifically for the Smith Learning Theater.
Video feeds throughout the Theater are processed and 'captioned' with Goodwin's program, then visualized through a projection. The output of the 'word.camera' is a poetic narrative that describes the footage being taken in real time using the lexicon of TCR as constructed by the neural network. No recordings other than the language are taken.
* Image taken from user testing