AI Paper Atlas

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

"Propose, Assess, Search: Harnessing LLMs for Goal-Oriented Planning in Instructional Videos"

Md Mohaiminul Islam, Tushar Nagarajan, Huiyu Wang, FU-JEN CHU, Kris Kitani, Gedas Bertasius, Xitong Yang

ECCV 2024 · Vision-language and multimodal › Robot action and navigation · approach: Pretraining and data augmentation for VLN · 7 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 propose-assess-search planning on top of LLM prompting (zero-shot, COIN VPA T=3)Llama-2-70B LLM Baseline · COIN VPA (zero-shot, T=3)success rate: 2.4 -> 15.3Improved
add propose-assess-search planning (few-shot, COIN VPA T=3)Llama-2-70B LLM Baseline · COIN VPA (few-shot, T=3)success rate: 10.2 -> 21.8Improved
add assessment and search over the multi-candidate LLM agent baselineLLM Agent Baseline · COIN VPA (zero-shot, T=3)success rate: 3.1 -> 15.3Improved
few-shot VidAssist vs fully supervised VLaMPVidAssist · COIN VPA T=4success rate: 9.0 -> 13.8Improved
few-shot VidAssist vs fully supervised LFP on COIN PPVidAssist · COIN PP T=3success rate: 30.64 -> 29.20Hurt
add LLM-based partial plan evaluation value function (ablation)VidAssist · COIN few-shot, VPA T=3success rate: 16.96 -> 21.08Improved
use partial plan evaluation alone vs generation score alone (ablation)VidAssist value functions · COIN few-shot, VPA T=3success rate: 11.60 -> 16.10Improved
add few-shot task graph value function (ablation)VidAssist · COIN few-shot, PP T=4success rate: 19.03 -> 20.78Improved
scale LLM from Llama-2-7B to Llama-2-70B (ablation)VidAssist · COIN few-shot, VPA T=4success rate: 11.71 -> 13.80Improved
increase in-context examples from 0 to 3 (ablation)VidAssist · COIN VPA T=3success rate: 15.30 -> 21.08Improved
use ground-truth action history instead of predicted history (ablation)VidAssist · COIN VPAsuccess rate: T=3: 21.08 -> 41.33Improved

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