Systematic Multi-Agent Vision-and-Language Navigation: Formulation, Benchmark, and Method

作者：Yunzhe Xu, Zhe Liu · 单位：general intelligence in the real world, as team collaboration, include collaborative exploration [11], scene understanding, The authors are with the School of Automation and Intelligent Sensing, Shanghai Jiao Tong University, Shanghai 200240, China (e-mail:, we present Topological Memory-based Collaborative Nav- · 会议/期刊：arXiv preprint · 方向：cs.CV · 发布日期：2026-09-30