Enhanced Human-Machine Interaction by Combining Proximity Sensing with Global Perception

TL;DR

This paper presents an optical system that combines proximity sensing and global perception to enhance human-robot interaction in large workspaces, enabling accurate 3D pose estimation from a single panoramic camera.

Contribution

It introduces a novel optical system that merges posture context with proximity perception, improving long-distance human pose estimation in collaborative robotics.

Findings

System predicts metric 3D human pose over large fields of view.

Combines proximity sensors with global perception to reduce occlusions.

Demonstrates effectiveness in multi-human, multi-robot scenarios.

Abstract

The raise of collaborative robotics has led to wide range of sensor technologies to detect human-machine interactions: at short distances, proximity sensors detect nontactile gestures virtually occlusion-free, while at medium distances, active depth sensors are frequently used to infer human intentions. We describe an optical system for large workspaces to capture human pose based on a single panoramic color camera. Despite the two-dimensional input, our system is able to predict metric 3D pose information over larger field of views than would be possible with active depth measurement cameras. We merge posture context with proximity perception to reduce occlusions and improve accuracy at long distances. We demonstrate the capabilities of our system in two use cases involving multiple humans and robots.