Diver Identification Using Anthropometric Data Ratios for Underwater Multi-Human-Robot Collaboration

TL;DR

This paper introduces a novel diver identification algorithm for underwater robots that uses anthropometric data and a specialized embedding network, achieving high accuracy in controlled experiments.

Contribution

The paper presents a new diver identification method leveraging anthropometric features and an embedding network, improving robustness and accuracy over existing approaches.

Findings

High accuracy in controlled-water experiments

Robust features invariant to distance and lighting changes

Effective embedding network for diver classification

Abstract

Recent advances in efficient design, perception algorithms, and computing hardware have made it possible to create improved human-robot interaction (HRI) capabilities for autonomous underwater vehicles (AUVs). To conduct secure missions as underwater human-robot teams, AUVs require the ability to accurately identify divers. However, this remains an open problem due to divers' challenging visual features, mainly caused by similar-looking scuba gear. In this paper, we present a novel algorithm that can perform diver identification using either pre-trained models or models trained during deployment. We exploit anthropometric data obtained from diver pose estimates to generate robust features that are invariant to changes in distance and photometric conditions. We also propose an embedding network that maximizes inter-class distances in the feature space and minimizes those for the intra-class features, which significantly improves classification performance. Furthermore, we present an end-to-end diver identification framework that operates on an AUV and evaluate the accuracy of the proposed algorithm. Quantitative results in controlled-water experiments show that our algorithm achieves a high level of accuracy in diver identification.