AI Paper Atlas

Research map / Vision-language and multimodal / Human and embodied modeling

Grounded Entity-Landmark Adaptive Pre-Training for Vision-and-Language Navigation

Yibo Cui, Liang Xie, Yakun Zhang, Meishan Zhang, Ye Yan, Erwei Yin

ICCV 2023 · Vision-language and multimodal › Human and embodied modeling · approach: Scene graph generation · 29 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add GELA adaptive pre-training to HAMT (none aug)HAMT · R2R val unseenSR: 65.7 -> 69.2Improved
add GELA adaptive pre-training to HAMTHAMT · R2R val unseen (NE)NE: 2.29 -> 3.33Hurt
add EPP alone (ablation)HAMT · R2R val unseen (none aug)SR: 65.7 -> 65.8No clear effect
add LBP alone (ablation)HAMT · R2R val unseen (none aug)SR: 65.7 -> 65.6No clear effect
add ELSA alone (ablation)HAMT · R2R val unseen (none aug)SR: 65.7 -> 66.3Improved
combine EPP and LBP (ablation)HAMT · R2R val unseen (none aug)SR: 65.8 -> 67.0Improved
add ELSA on top of EPP+LBP (ablation)GELA · R2R val unseen (none aug)SR: 67.0 -> 69.2Improved
use scene entities only (ablation)GELA · R2R val unseen (none aug)SR: 65.7 -> 66.7Improved
use both object and scene entities versus object only (ablation)GELA · R2R val unseen (none aug)SR: 66.8 -> 69.2Improved
GELA vs HAMT on CVDN val seenHAMT · CVDN val seenGoal Progress (m): 6.91 -> 8.57Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy