Edge-assisted collaborative image recognition for augmented reality: Demo abstract

Published

Conference Paper

© 2019 Authors. Mobile Augmented Reality (AR), which overlays digital information with real-world scenes surrounding a user, provides an enhanced mode of interaction with the ambient world. Contextual AR applications rely on image recognition to identify objects in the view of the mobile device. In practice, due to image distortions and device resource constraints, achieving high performance image recognition for AR is challenging. Recent advances in edge computing offer opportunities for designing collaborative image recognition frameworks for AR. In this demonstration, we present CollabAR, an edge-assisted collaborative image recognition framework. CollabAR allows AR devices that are facing the same scene to collaborate on the recognition task. Demo participants develop an intuition for different image distortions and their impact on image recognition accuracy. We showcase how heterogeneous images taken by different users can be aggregated to improve recognition accuracy and provide a better user experience in AR.

Full Text

Duke Authors

Cited Authors

  • Stojkovic, J; Liu, Z; Lan, G; Joe-Wong, C; Gorlatova, M

Published Date

  • November 10, 2019

Published In

  • Sensys 2019 Proceedings of the 17th Conference on Embedded Networked Sensor Systems

Start / End Page

  • 394 - 395

International Standard Book Number 13 (ISBN-13)

  • 9781450369503

Digital Object Identifier (DOI)

  • 10.1145/3356250.3361944

Citation Source

  • Scopus