Reproduce “SHARE: Towards Head-Mounted AR with User-Centric SLAM in Shared Human-Robot Workspaces” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Tianyuan Du, Tianyi Hu, Hanting Ye, Maria Gorlatova

SHARE is a user-centric multi-agent SLAM system for human-robot collaboration that prioritizes augmented reality (AR) user experience. It achieves an average motion-to-photon latency of 13.22 ms for AR users, a 43.3% reduction over baselines, while maintaining robot tracking accuracy below 2 cm ATE across diverse environments.

Key insights

  • Prioritizing the stringent latency demands of human AR users while leveraging the higher latency tolerance of robots in HRC scenarios significantly enhances overall system performance and user experience.
  • Adaptive resource allocation, driven by real-time accuracy estimation and a continuous feedback loop via a PID controller, is essential for maintaining optimal QoE for heterogeneous agents in dynamic multi-agent SLAM.
  • Computational efficiency in multi-agent SLAM within small, shared workspaces can be effectively improved by strategically identifying and reducing redundant map merging through lightweight visual overlap assessment.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.