AI Paper Atlas

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

VLBiMan: Vision-Language Anchored One-Shot Demonstration Enables Generalizable Bimanual Robotic Manipulation

Huayi Zhou, Kui Jia

ICLR 2026 · Vision-language and multimodal › Robot action and navigation · approach: Memory and subgoal-based manipulation

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 Florence-2 + SAM2 instead of SAM+DINOv2 for perception (ablation)VLBiMan · Six basic tasks, novel instances, with interferenceaverage success rate (%): 35.8% -> 59.2%Improved
use the proposed initial grasp alignment instead of AnyGrasp (ablation)VLBiMan · Six basic tasks, novel instances, with interferenceaverage success rate (%): 31.7% -> 59.2%Improved
add progressive IK refinement (ablation)VLBiMan · Six basic tasks, novel instances, with interferenceaverage success rate (%): 29.2% -> 59.2%Improved
add dynamic collision compensation (ablation)VLBiMan · Six basic tasks, novel instances, with interferenceaverage success rate (%): 34.2% -> 59.2%Improved
inject oracle initial grasp label into ReKep (ReKep+)ReKep · Six basic tasks, same objects, no interferenceaverage success rate (%): 43.3% -> 63.3%Improved
add external dynamic interference during executionVLBiMan · Six basic tasks, novel instancesaverage success rate (%): 78.0% -> 59.3%Hurt
replace same objects with novel category-level instancesVLBiMan · Six basic tasks, no interferenceaverage success rate (%): 85.3% -> 78.0%Hurt
move from basic tasks to long-horizon tasksVLBiMan · same objects, no interferenceaverage success rate (%): 85.3% -> 52.0%Hurt

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