AI Paper Atlas

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

ARNOLD: A Benchmark for Language-Grounded Task Learning with Continuous States in Realistic 3D Scenes

Ran Gong, Jiangyong Huang, Yizhou Zhao, Haoran Geng, Xiaofeng Gao, Qingyang Wu, Wensi Ai, Ziheng Zhou, Demetri Terzopoulos, Song-Chun Zhu, Baoxiong Jia, Siyuan Huang

ICCV 2023 · Vision-language and multimodal › Robot action and navigation · approach: Multimodal prompt-driven manipulation · 53 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 language instruction (ablation)PerAct · ARNOLD Testaverage success rate (%): 17.67 -> 42.47Improved
add explicit state-value supervision head (ablation)PerAct · ARNOLD Testaverage success rate (%): 42.47 -> 43.36No clear effect
add explicit state-value supervision head (ablation)PerAct · ARNOLD Novel Stateaverage success rate (%): 4.49 -> 4.10No clear effect
add explicit state-value supervision head (ablation)PerAct · ARNOLD Any Stateaverage success rate (%): 16.36 -> 19.97Improved
add explicit state-value supervision head (ablation)PerAct (MT) · ARNOLD Testaverage success rate (%): 29.14 -> 31.00Improved
add explicit state-value supervision head, given first-phase ground truth (ablation)PerAct · ARNOLD Test (first-phase ground truth)average success rate (%): 55.24 -> 60.22Improved
add first-phase ground truth (grasp positions) (ablation)PerAct · ARNOLD Testaverage success rate (%): 42.47 -> 55.24Improved
train multi-task instead of single-task (ablation)PerAct · ARNOLD Testaverage success rate (%): 42.47 -> 29.14Hurt
replace CLIP language encoder with T5-base (ablation)PerAct · ARNOLD OpenDrawersuccess rate (%)Improved
use voxel-grid PerAct instead of top-down image 6D-CLIPort6D-CLIPort · ARNOLD Testaverage success rate (%): 0.84 -> 42.47Improved
use PerAct instead of single-view regression BC-Lang-ViTBC-Lang · ARNOLD OpenDrawer Testsuccess rate (%): 5.88 -> 31.09Improved
evaluate on novel goal states instead of seen onesPerAct · ARNOLD Novel Stateaverage success rate (%): 42.47 -> 4.49Hurt

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