Object-Independent Human-to-Robot Handovers using Real Time Robotic Vision

TL;DR

This paper introduces a real-time, object-independent human-to-robot handover system that uses robotic vision and manipulation to ensure safe and efficient exchanges without relying on external sensors.

Contribution

It presents a novel approach combining a generic object detector, fast grasp selection, and human safety modules using only a single RGB-D camera for real-time handovers.

Findings

81.9% success rate in object handover trials

Real-time perception modules enable safe and efficient handovers

Object-independent approach adaptable to various objects

Abstract

We present an approach for safe and object-independent human-to-robot handovers using real time robotic vision and manipulation. We aim for general applicability with a generic object detector, a fast grasp selection algorithm and by using a single gripper-mounted RGB-D camera, hence not relying on external sensors. The robot is controlled via visual servoing towards the object of interest. Putting a high emphasis on safety, we use two perception modules: human body part segmentation and hand/finger segmentation. Pixels that are deemed to belong to the human are filtered out from candidate grasp poses, hence ensuring that the robot safely picks the object without colliding with the human partner. The grasp selection and perception modules run concurrently in real-time, which allows monitoring of the progress. In experiments with 13 objects, the robot was able to successfully take the object from the human in 81.9% of the trials.