Reproduce “Air-Ground Collaborative Vision-and-Language Navigation via Shared Bird's-Eye Maps” 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
Shuning Zhang, Liang Li, Yunheng Wang, Tao Wang, Yihang Kang, Renjing Xu
AGC-VLN coordinates a drone’s overhead view with a road vehicle navigating to a static target truck. The drone marks both on an overhead image; the vehicle plans a road route while the drone flies toward the target. Frozen vision-language models guide planning, and geometry and controllers execute movement. In 100 CARLA-Air episodes, joint success is 77.0%, 27 points above the weaker agent, on this benchmark.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.