eCAR: edge-assisted Collaborative Augmented Reality Framework

TL;DR

eCAR tackles high traffic and latency in multi-user Collaborative AR by introducing an edge-assisted framework that continuously aligns device coordinate systems and extends a co-visibility graph for spatial-temporal consistency. It introduces a graph-grid-data structure and an edge-server–driven coordination scheme to synchronize SLAM maps and virtual objects across devices with reduced network load. The authors demonstrate up to 370% traffic reduction and up to 62% latency reduction (up to 20 devices) while preserving localization accuracy, validating the approach on public datasets and real indoor setups. This work enables scalable, low-latency, multi-user AR experiences in large indoor environments like museums and shopping centers by leveraging edge computing and grid-based object sharing.

Abstract

We propose a novel edge-assisted multi-user collaborative augmented reality framework in a large indoor environment. In Collaborative Augmented Reality, data communication that synchronizes virtual objects has large network traffic and high network latency. Due to drift, CAR applications without continuous data communication for coordinate system alignment have virtual object inconsistency. In addition, synchronization messages for online virtual object updates have high latency as the number of collaborative devices increases. To solve this problem, we implement the CAR framework, called eCAR, which utilizes edge computing to continuously match the device's coordinate system with less network traffic. Furthermore, we extend the co-visibility graph of the edge server to maintain virtual object spatial-temporal consistency in neighboring devices by synchronizing a local graph. We evaluate the system quantitatively and qualitatively in the public dataset and a physical indoor environment. eCAR communicates data for coordinate system alignment between the edge server and devices with less network traffic and latency. In addition, collaborative augmented reality synchronization algorithms quickly and accurately host and resolve virtual objects. The proposed system continuously aligns coordinate systems to multiple devices in a large indoor environment and shares augmented reality content. Through our system, users interact with virtual objects and share augmented reality experiences with neighboring users.

Figures (12)

• Figure 1: System Overview.
• Figure 2: Proposed Graph-Grid-Data Structure. VO is a virtual object. KF is a keyframe, MP is a map point. Observations connect KF and MP. Graph connects grid cells observed in the viewpoint of the keyframe using structure planes and the frame's pose. By user interaction input, virtual objects are connected with a grid cell in for synchronization.
• Figure 3: Local Graph Update in the device. A device manages a local graph using the queue of keyframes. The device selects the image sent to the server as the key frame for coordinate alignment.
• Figure 4: Example of Virtual Object Manipulation.
• Figure 5: Example of Asynchronous Communication. The sky blue square is a graph synchronization latency. The upper column is the data communication interval for matching in the device, and the lower column is the frame interval. In the example, the matching interval is assumed to be four frames. The red rectangle is the virtual object registration interaction latency, and the green rectangle is the virtual object update interaction latency. A blue borderline of the square is a status message visualized on the device $d_a$. eCAR synchronizes virtual objects by each device in graph update.