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

TL;DR

VLBiMan addresses the challenge of generalizable dual-arm manipulation from minimal human input by grounding one-shot demonstrations with vision-language anchors to reusable atomic skills. It decomposes tasks into invariant primitives and adapts variable components through VLM-based scene grounding and geometry-aware anchors, then composes trajectories with kinematic refinements. The approach demonstrates cross-embodiment transfer, robustness to clutter and disturbances, and effective long-horizon skill composition, reducing the need for extensive demonstrations. This work bridges human priors and foundation-model grounding to enable practical, versatile bimanual manipulation in unstructured environments. Across ten tasks and cross-embodiment scenarios, VLBiMan shows strong generalization, robustness, and efficiency gains over strong baselines.

Abstract

Achieving generalizable bimanual manipulation requires systems that can learn efficiently from minimal human input while adapting to real-world uncertainties and diverse embodiments. Existing approaches face a dilemma: imitation policy learning demands extensive demonstrations to cover task variations, while modular methods often lack flexibility in dynamic scenes. We introduce VLBiMan, a framework that derives reusable skills from a single human example through task-aware decomposition, preserving invariant primitives as anchors while dynamically adapting adjustable components via vision-language grounding. This adaptation mechanism resolves scene ambiguities caused by background changes, object repositioning, or visual clutter without policy retraining, leveraging semantic parsing and geometric feasibility constraints. Moreover, the system inherits human-like hybrid control capabilities, enabling mixed synchronous and asynchronous use of both arms. Extensive experiments validate VLBiMan across tool-use and multi-object tasks, demonstrating: (1) a drastic reduction in demonstration requirements compared to imitation baselines, (2) compositional generalization through atomic skill splicing for long-horizon tasks, (3) robustness to novel but semantically similar objects and external disturbances, and (4) strong cross-embodiment transfer, showing that skills learned from human demonstrations can be instantiated on different robotic platforms without retraining. By bridging human priors with vision-language anchored adaptation, our work takes a step toward practical and versatile dual-arm manipulation in unstructured settings.

Figures (15)

• Figure 1: Left: Taking pouring water as an example, we sketch the entire process of VLBiMan based on the one-shot demonstration. Right: VLBiMan can achieve generalizable bimanual manipulation on a variety of complex contact-rich tasks without retraining, robustly coping with diverse scenarios.
• Figure 2: Framework of Vision-Language Anchored Bimanual Manipulation (VLBiMan). Taking the pouring water as an example, the paradigm consists of three stages (e.g., decomposition, adaptation, and composition) based on a given demonstration. VLBiMan can achieve generalization of unseen spatial placements and category-level new instances under the same task.
• Figure 3: Illustrations of representative points for manipulated objects in three tasks: pouring (left), reorient+unscrew (middle) and tool-use:spoon (right). These points will be used to calculate the change in object position and orientation (not always required).
• Figure 4: Manipulated object assets involved in each task, and the fixed-base dual-arm platform.
• Figure 5: Visualization of ten tasks executed on real robots. They are designed to validate different aspects, including (a) six dual-arm primary skills, (b) combination of basic skills for two long-horizon tasks, and (c) exploration of two multi-stage tool-use tasks.