Reproduce “HOP: History-and-Order Aware Pre-training 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

Yanyuan Qiao, Yuankai Qi, Yicong Hong, Zheng Yu, Peng Wang, Qi Wu

HOP pretrains vision-and-language navigation agents to use visual history and observation order when predicting actions. It combines cross-modal matching with fine- and coarse-grained order-recovery tasks and history-conditioned action prediction. HOP reports 59 SPL on R2R test-unseen, versus 57 for RecBERT and AirBERT, but its model also uses processed BnB data, so the comparison does not isolate the objectives.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.