LOVE ACTUALLY ... COMMON OBJECTS ... DANJOU ... THE BLIND SCANNER ... NO STOP CLASSROOM ... A NATION IS SOMETHING THIN ... DISPLACEMENT AND RESISTANCE ... HARD FOREST ... SHELL CONSTRUCTION ... SHELF LIFE ... ALGORITHMS OF THE VERNACULAR ... THE EVERYWHERE STORE ... NO STOP SHOPPING ... THE END OF BUILDINGS ... THE ROOKERY ... THE CROOKED AND THE STRAIGHT ... BLACKBOX / BORDERLANDS, PART I ... BLACKBOX / BORDERLANDS, PART II ... THE VIRTUAL CITY ... BUILDING CODE ... PENDULUM
Common Objects is a series of ongoing experiments in human-AI collaborative problem solving. It is concerned with the “problems of assembly” often encoutered by designers across different disciplines, in which pre-existing, well-defined objects must be collected and arranged in order to satisfy complex, open-ended needs and desires.

Using COCO, a large open-source image dataset, and YOLOv4, a neural network for object detection and classification, relationships between objects in a scene are detected, analyzed, and projected back in real time, allowing a participant to react to the analysis spontaneously and improvisationally.

SPRING 2021