The article presents Multi S-Graphs, a novel decentralized Collaborative Simultaneous Localization and Mapping (CSLAM) system. Traditional CSLAM techniques are limited by the extensive data required for transmission and a lack of deep environmental understanding. Multi S-Graphs address these issues by utilizing high-level semantic-relational information. This new approach improves the accuracy and robustness of the CSLAM system while reducing the amount of data exchanged between robots. The system has been tested in both simulated and real environments.
Publication date: 12 Jan 2024
Project Page: https://github.com/snt-arg/multi_s_graphs_docker
Paper: https://arxiv.org/pdf/2401.05152
