AI Paper Atlas

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

MobileManiBench: Simplifying Model Verification for Mobile Manipulation

Wenbo Wang, Fangyun Wei, Qixiu Li, Xi Chen, Yaobo Liang, Chang Xu, Jiaolong Yang, Baining Guo

ECCV 2026 · Vision-language and multimodal › Robot action and navigation · approach: Memory and subgoal-based manipulation · 1 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 depth to multi-view RGB input (ablation)MobileManiVLA · MobileManiBench G1 ablation subsetsuccess rate (%): 14.9 -> 28.2Improved
add wrist view to head-view RGB-D (ablation)MobileManiVLA · MobileManiBench G1 ablation subsetsuccess rate (%): 14.1 -> 28.2Improved
add depth to head-view RGB (ablation)MobileManiVLA · MobileManiBench G1 ablation subsetsuccess rate (%): 7.9 -> 14.1Improved
add wrist view to head-view RGB (ablation)MobileManiVLA · MobileManiBench G1 ablation subsetsuccess rate (%): 7.9 -> 14.9Improved
add hand pose state input (ablation)MobileManiVLA · MobileManiBench G1 ablation subsetsuccess rate (%): 22.4 -> 28.2Improved
add object grasp point state (ablation)MobileManiVLA · MobileManiBench G1 ablation subsetsuccess rate (%): 28.2 -> 32.3Improved
add goal point state (ablation)MobileManiVLA · MobileManiBench G1 ablation subsetsuccess rate (%): 32.3 -> 36.6Improved
add hand pose state input (ablation)MobileManiVLA · MobileManiBench G1 ablation subset (cart)success rate (%): 28.1 -> 23.5Hurt
test on seen scenes instead of unseen scenes (unseen objects)MobileManiVLA · MobileManiBench G1 ablation subsetsuccess rate (%): 28.2 -> 39.2Improved
test on seen objects instead of unseen objects (unseen scenes)MobileManiVLA · MobileManiBench G1 ablation subsetsuccess rate (%): 28.2 -> 51.3Improved
use MobileManiVLA instead of pi0.5pi0.5 · MobileManiBench G1 ablation subsetsuccess rate (%): 18.8 -> 28.2Improved
use MobileManiVLA instead of OpenVLAOpenVLA · MobileManiBench G1 ablation subsetsuccess rate (%): 4.5 -> 28.2Improved

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