Date of Award
Spring 5-9-2026
Document Type
Honors Thesis
Department/Major
Biomedical Engineering
First Advisor
Dr. Lisa MacFadden
Second Advisor
Dr. Ryan Rykhus
Third Advisor
N/A
Keywords
RF-Sensing, Wi-Fi-Sensing, Pose Estimation, Motion Capture, Emerging Technologies, Healthcare Technologies, Policy Gaps, Technical Survey, Surveillance, Intelligence Analysis
Abstract
Wi-Fi-based sensing has emerged as a promising passive monitoring method, exploiting Channel State Information from commodity wireless hardware to infer human presence, motion, and physiological signals without cameras or wearable devices. This thesis presents a structured technical survey and critical evaluation of Wi-Fi-based human pose estimation, examining the signal processing foundations, machine learning architectures, and evaluation practices that define the current state of the field. The survey traces the field's development from coarse activity recognition to fine-grained skeletal inference, critically evaluating architectural trade-offs and identifying evaluation limitations that complicate assessment of real-world readiness. Healthcare and security applications are examined as primary evaluative lenses, finding that current systems are best suited to coarse-grained inference tasks and that deployment readiness remains constrained by environmental sensitivity and the absence of prospective operational validation. The ethical and regulatory implications of passive RF sensing are examined, identifying gaps around consent, jurisdictional coverage, and liability. The thesis concludes with a proposed application-driven design framework and a set of technical and policy priorities necessary for Wi-Fi-based sensing to transition responsibly from research demonstration to practical deployment.
Recommended Citation
Johnson, Landon D., "Seeing Through Walls: Applications and Ethical Implications of Wi-Fi-Based Human Pose Estimation" (2026). Honors Thesis. 428.
https://red.library.usd.edu/honors-thesis/428