IAR: An Intelligent Augmented Reality Framework

Department of Computer Science and Engineering

Project Abstract

The field of Artificial Intelligence (AI) studies approaches to utilize computers to perform human tasks. On the other hand, Augmented Reality (AR) is a technology field where computer graphics is harnessed in a user’s view of the real world. On their own, each individual computer science field mentioned above provides powerful aid and utility in improving everyday life. In this project, our ultimate goal is to seek out a framework (i.e., a general-purpose software that can be customized to implement different applications) that would yield the optimal marriage of these two aforementioned fields. Our framework, dubbed IAR (Intelligent Augmented Reality), consists of modules for data preprocessing, data management and organization, data modeling (with machine learning and deep learning models), and data visualization. 

There is a plethora of applications that would see great benefit from the integration of AI and AR using IAR. For demonstration purposes, we have implemented three specific applications from diverse areas based on the IAR framework: 1) a tool that enables intelligent lane detection on road networks under adverse weather conditions and/or for visually impaired, 2) a tool that serves as therapeutic aid for patients with strabismus, and 3) a tool that allows for automated classification of garbage to educate the user on the differences between recyclable and non-recyclable trash.

Project presentation

Project poster

IAR: Intelligent augmented reality framework