Back to search

University of North Texas

Remote Human-in-the-Loop (RHIL) for Real-time Artificial Intelligence (AI) Systems

Abstract

dc:description

This dissertation investigates and discusses the existing methods and technologies for real-time assessment of physical activities, their feedback mechanisms, and their limitations. These limitations are addressed using modern artificial intelligence techniques in collaboration with a remote human expert. This research study focuses on physiotherapy, physical exercise, and cardiopulmonary resuscitation (CPR); the proposed frameworks can also be adapted to other domains. AI-based human pose estimation is the key computer vision technique utilized to develop most of the proposed frameworks. In addition, various machine learning algorithms, including reinforcement learning and federated learning, as well as computer vision techniques such as optical flow and noise-cancellation methods, were employed depending on the specific requirements of the application. The contributions of this research include: (1) A novel collaborative and adaptive feedback mechanism for physical exercises, (2) A novel metric for evaluating the difficulty associated with a repetition of a physical exercise, (3) A non-intrusive, markerless, and infrastructure-independent framework that provides prompt high-quality feedback for CPR using optical sensors, (4) A non-intrusive, markerless, and infrastructure-independent framework that provides prompt feedback for ventilation in CPR training, and (5) A blockchain-based authorization framework for remote collaboration systems. The proposed frameworks are designed with accessibility in mind, making them both cost-effective and user-friendly. They were validated through human subject experiments and statistical analysis to evaluate effectiveness. These AI-powered frameworks provide an innovative approach to effective learning, ultimately improving patient care.

Degree

thesis:*
Grantor dc:publisher
University of North Texas
Year dc:date
2026

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Ranasinghe Arachchilage, Ishan, 1986-
Contributors dc:contributor
  • Dantu, Ramanamurthy
  • Thompson, Mark
  • Albert, Mark V.
  • Morozov, Kirill
  • Hauge, Janice A.

Subjects

dc:subject × 3

Rights

dc:rights
Statement dc:rights
  • Public
  • Ranasinghe Arachchilage, Ishan, 1986-
  • Copyright
  • Copyright is held by the author, unless otherwise noted. All rights Reserved.
Language dc:language
English

Identifiers

dc:identifier.*
Identifier
local-cont-no: submission_4714
ark: ark:/67531/metadc2588600

Chain of custody

source
Harvested from
University of North Texas
Base URL
digital.library.unt.edu/oai/
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Ranasinghe Arachchilage, Ishan, 1986-. Remote Human-in-the-Loop (RHIL) for Real-time Artificial Intelligence (AI) Systems. University of North Texas, 2026. https://digital.library.unt.edu/ark:/67531/metadc2588600/