Beholder

beholder Banner

Beholder is a Computer Vision (CV) toolkit for building tangible controllers for interactive computer systems. It facilitates makers to build physical inputs for a computer using the ArUco CV markers using a visual, node-based programming interface. One can connect behaviors of CV markers in the physical world to any software that responds to keyboard input. Each marker has properties that can be detected by a CV system, such as presence, position, rotation, and tangible inputs can be designed to manipulate these properties and trigger a response (e.g., keyboard event). The Beholder editor runs as a standalone desktop application on the computer that receives and processes the camera feed from a webcam. It provides a visual node-based programming environment for makers to create logical relationships between CV marker behavior and keyboard events. The editor offers debugging in real-time with a live camera view that highlights detected markers and their properties. Ìý

The Beholder project explores a CV-driven approach to physical computing, a traditionally electronics-based technique for enabling both young learners and interaction designers to build physical computational artifacts, like controllers. Unlike electronics, Beholder and CV offer a low-barrier, low-cost, and no-code method to detect physical input to computing systems. We conducted design studios and workshops with students to reveal how fiducial markers may be used too sense interactions like rotation, pressing, shaking, lighting, etc., otherwise sensed with electronic components like potentiometers, buttons, accelerometers, or light sensors. Beholder provides a platform for engaging young learners and non-experts to build and prototype physical computing artifacts using everyday materials like paper and cardboard. Over the years, we have utilized the Beholder editor in workshops, hackathons, and week-long projects to engage undergraduate and high school students in ATLAS, NCKU Taiwan, and DSST schools with CV-based physical computing.

ACME LabÌý

Two people holding up too cardboard controllers that resemble guitar heads

Ìý

A cardboard skateboard controller

Ìý

Two people playing with racket shaped cardboard controllers in front of a laptop

Ìý

Person holding a skateboard shaped cardboard controller

Ìý

Person coloring a cardboard racket

Associated Researcher

Publications

Peter Gyory, S. Sandra Bae, Ruhan Yang, Ellen Yi-Luen Do, and Clement Zheng. 2023. ". In: Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems (CHI '23).

Peter Gyory, Krithik Ranjan, Zhen Zhou Yong, Clement Zheng, and Ellen Yi-Luen Do. 2022. ". In: SIGGRAPH Asia 2022 Emerging Technologies(SA '22). Ìý

Peter Gyory. 2022. "". In: Creativity and Cognition (C&C '22).(June 20-23, 2022—Venice, Italy).

Beholder News