2020-RAMA-69137 | |
Researchers at Purdue University have developed a new tool for augmented reality in enhance machine awareness to human interaction, known as CAPturAR. Currently, machines recognize human interaction in pre-defined contexts but have difficulty recognizing day-to-day personal interactions. Purdue researchers meet this challenge through a helmet-like device with a multi-camera setup that allows machines to observe human activity actively and rapidly author computer programs. Common types of human activity were tested with a prototype device including a participant survey twelve unique users opening a pill bottle at the same time daily which then took a computer 0.65 seconds to replicate and at a position that was off by 3.69 cm on average. In addition, computers were able to learn sequential tasks from participants such as repairing a bicycle wheel. Advantages: -Accurate -Rapid Authoring -Repeat Measurements Potential Applications: -Machine Learning -Robotics Technology Validation: Participant study Recent Publication Convergence Design Lab Purdue University 2020 UIST 33rd ACM User Interface Software and Technology Symposium engineering.purdue.edu/cdesign/wp/ |
|
|
|
Jun 30, 2021
Utility-Gov. Funding
United States
(None)
(None)
Jun 30, 2020
Provisional-Gov. Funding
United States
(None)
(None)
|
|
Purdue Office of Technology Commercialization The Convergence Center 101 Foundry Drive, Suite 2500 West Lafayette, IN 47906 Phone: (765) 588-3475 Fax: (765) 463-3486 Email: otcip@prf.org |