Date Complete

Spring 2026

Document Type

Poster

Department

Engineering & Computing Sciences

First Advisor

Dr. Robin Ghosh

Abstract

This project investigates the feasibility of real-time full-body movement classification using electroencephalography (EEG) integrated with virtual reality (VR) technologies. The primary objective is to develop and evaluate machine learning models for predicting human body movements using EEG data alone, with the long-term goal of reducing or eliminating reliance on wearable motion trackers. Currently, several machine learning algorithms have been tested, but classification accuracy remains modest, indicating the complexity of the task. Ongoing work focuses on optimizing preprocessing, feature selection, and model architectures to improve performance. The system architecture combines synchronized neural and motion data collected within a VR environment. EEG signals are acquired using the Neurosity Crown, an 8-channel EEG headset, while ground-truth body-movement data are captured using eight Somatic VR inertial measurement units (IMUs) as full-body trackers. A Meta Quest Pro VR headset provides an immersive testing environment and supports additional modalities such as facial tracking. EEG and motion data are temporally aligned and preprocessed to reduce noise and artifacts before being used to train machine learning models that map neural activity to specific body movements. The system is designed to operate in real time, enabling continuous decoding of movement intent from EEG signals. By leveraging VR full-body tracking as a training reference, the model aims to learn neural representations of motor activity. The architecture is extensible, enabling future integration of complementary data sources, such as eye-tracking and muscle-based sensors, to improve robustness and classification accuracy. The outcomes of this work have implications for hands-free VR interaction, accessibility technologies, and non-invasive neuroprosthetic systems that interpret motor intent directly from brain activity. More broadly, this project contributes to understanding the relationship between neural signals and complex body movements, supporting future advances in neurotechnology and human-computer interaction.

Download

Share

COinS