Reproduce “Seeing Less in a Sparse World: Cross-View Geolocalization from Limited Field-of-View Observations” 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
Fahimul Aleem, Shruti Vyas
GeoQueryNet retrieves matching satellite images for ground-level camera queries. It uses learned query tokens to attend first to ground features and then satellite features, with LoRA adaptation. On CVW500k, it achieves 15.19% Recall@1, compared with QDFL’s 8.19%. However, coverage is sparse and uneven, and inference costs 22.04 ms and 527.38 GFLOPs per image pair.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.