AI Paper Atlas

Research map / Vision-language and multimodal / Robot action and navigation

NavGPT-2: Unleashing Navigational Reasoning Capability for Large Vision-Language Models

Gengze Zhou, Yicong Hong, Zun Wang, Xin Eric Wang, Qi Wu

ECCV 2024 · Vision-language and multimodal › Robot action and navigation · approach: Reasoning-enhanced vision-language navigation · 88 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
use LLM latents (frozen VLM + Q-former) as visual-linguistic representation instead of LXMERT-style transformerDUET global-action policy · R2R Test UnseenSR: 69 -> 71 (FlanT5-XL w/ PREVALENT)Improved
use FlanT5-XXL w/ PREVALENT NavGPT-2 over NaviLLMNavGPT-2 · R2R Test UnseenSR: 68 -> 72Improved
remove the navigation-specific policy network (ablation)NavGPT-2 FlanT5-XL · R2R Val UnseenSR: 67.52 -> 21.46Hurt
pretrain Q-former on navigational reasoning data (ablation)NavGPT-2 FlanT5-XL · R2R Val UnseenSR: 66.75 -> 67.52Improved
scale LLM from FlanT5-XL to FlanT5-XXL (ablation)NavGPT-2 · R2R Val UnseenSR: 67.52 -> 71.31Improved
use decoder-only Vicuna instead of encoder-decoder FlanT5 (ablation)NavGPT-2 · R2R Val UnseenSR: 71.31 (FlanT5-XXL) -> 53.77 (Vicuna-7B), 48.28 (Vicuna-13B)Hurt
scale Vicuna from 7B to 13B (ablation)NavGPT-2 · R2R Val UnseenSR: 53.77 -> 48.28Hurt
use NavGPT-2 instead of DUET with limited training data (10%)NavGPT-2 FlanT5-XL · R2R Val Unseen (10% train)SR: 48.19 -> 52.02Improved
zero-shot transfer to unseen environments (HM3D) using LLM-space visual featuresNavGPT-2 FlanT5-XXL · HM3D zero-shotSR: 25.60 -> 47.20Improved
zero-shot transfer to RxR EnglishNavGPT-2 FlanT5-XXL · RxR-EN zero-shotSR: 25.07 -> 28.75Improved
swap in a stronger vision encoder for DUET (ablation)DUET global branch · R2R Val UnseenSRNo clear effect
remove local branch from DUET (ablation)DUET · R2R Test UnseenSR: 69 -> 67Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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