From Neck to Head: Bio-Impedance Sensing for Head Pose Estimation

TL;DR

NeckSense is a wearable bio-impedance system embedded in a necklace that accurately estimates head pose using deep learning, offering a line-of-sight-free alternative to vision-based methods with comparable accuracy.

Contribution

The paper introduces a novel bio-impedance wearable device and a deep learning framework that incorporates anatomical priors for robust head pose estimation.

Findings

Achieved a mean error of 25.9 mm in head pose estimation.

Validated on 7 participants with cross-validation, comparable to state-of-the-art vision methods.

Demonstrated robustness across various head movements.

Abstract

We present NeckSense, a novel wearable system for head pose tracking that leverages multi-channel bio-impedance sensing with soft, dry electrodes embedded in a lightweight, necklace-style form factor. NeckSense captures dynamic changes in tissue impedance around the neck, which are modulated by head rotations and subtle muscle activations. To robustly estimate head pose, we propose a deep learning framework that integrates anatomical priors, including joint constraints and natural head rotation ranges, into the loss function design. We validate NeckSense on 7 participants using the current SOTA pose estimation model as ground truth. Our system achieves a mean per-vertex error of 25.9 mm across various head movements with a leave-one-person-out cross-validation method, demonstrating that a compact, line-of-sight-free bio-impedance wearable can deliver head-tracking performance comparable to SOTA vision-based methods.