Reproduce “GridMM: Grid Memory Map for Vision-and-Language Navigation” 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
Zihan Wang, Xiangyang Li, Jiahao Yang, Yeqi Liu, Shuqiang Jiang
GridMM maps visual features from an agent’s past observations into a growing spatial memory, selecting within each cell the details most relevant to its navigation instruction. This combines spatial context with instruction-relevant visual clues. On the R2R-CE validation-unseen benchmark, it scores 49.05 SR and 40.99 SPL, above the topological-map and no-map entries; evidence is from simulation, and multi-floor navigation remains open.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.