Learning Humanoid End-Effector Control for Open-Vocabulary Visual Loco-Manipulation
2026-02-18 • Robotics
RoboticsComputer Vision and Pattern Recognition
AI summaryⓘ
The authors introduce HERO, a new method for humanoid robots to pick up and move everyday objects using visual inputs like RGB-D images. They improved control by combining classical robotics techniques with machine learning to create a more accurate way for the robot's hand to follow targets. This approach reduces tracking errors significantly and works well in many real-world places, such as offices and coffee shops. Their system uses large vision models to recognize objects and can handle items like mugs and toys on different heights of surfaces.
humanoid robotsend-effectorloco-manipulationRGB-D imagesimitation learninginverse kinematicsforward kinematicslarge vision modelstrajectory planningresidual tracking
Authors
Runpei Dong, Ziyan Li, Xialin He, Saurabh Gupta
Abstract
Visual loco-manipulation of arbitrary objects in the wild with humanoid robots requires accurate end-effector (EE) control and a generalizable understanding of the scene via visual inputs (e.g., RGB-D images). Existing approaches are based on real-world imitation learning and exhibit limited generalization due to the difficulty in collecting large-scale training datasets. This paper presents a new paradigm, HERO, for object loco-manipulation with humanoid robots that combines the strong generalization and open-vocabulary understanding of large vision models with strong control performance from simulated training. We achieve this by designing an accurate residual-aware EE tracking policy. This EE tracking policy combines classical robotics with machine learning. It uses a) inverse kinematics to convert residual end-effector targets into reference trajectories, b) a learned neural forward model for accurate forward kinematics, c) goal adjustment, and d) replanning. Together, these innovations help us cut down the end-effector tracking error by 3.2x. We use this accurate end-effector tracker to build a modular system for loco-manipulation, where we use open-vocabulary large vision models for strong visual generalization. Our system is able to operate in diverse real-world environments, from offices to coffee shops, where the robot is able to reliably manipulate various everyday objects (e.g., mugs, apples, toys) on surfaces ranging from 43cm to 92cm in height. Systematic modular and end-to-end tests in simulation and the real world demonstrate the effectiveness of our proposed design. We believe the advances in this paper can open up new ways of training humanoid robots to interact with daily objects.