Please use this identifier to cite or link to this item:
http://hdl.handle.net/11375/29390
Title: | KATNN: KAT Walk C Alternative Motion Capture Algorithm |
Other Titles: | KATNN: Motion Capture and Machine Learning to Predict Realistic Character Trajectory in a Virtual Game for the KAT Walk C |
Authors: | Matira, Kenneth |
Department: | Computing and Software |
Keywords: | unity;machine learning;virtual reality;motion capture;neural netwworks;locomotion;KATVR |
Publication Date: | Dec-2023 |
Publisher: | Department of Computing and Software, McMaster University |
Abstract: | In the world of Virtual Reality (VR), motion sickness, nausea, and disorientation remains a big concern for many users. The KAT Walk C is an omni-directional treadmill, which aims to convert human motion to virtual movement. This is intended to reduce the aforementioned concerns. However, the original KAT C algorithm of human locomotion has its limitations, where motion is frequently converted incorrectly. In this report, we will introduce an alternative input mechanism for the KAT Walk C, KATNN, which focuses on two primary objectives: allowing the user to move in multiple directions, and having the ability to register slower type motions. KATNN was created by the construction of modular neural networks. We will discuss steps to create the models, investigate current issues and potential solutions involving calibration and disorientation. Readers may optionally view the results by watching the following video: https://youtu.be/SbUXoQ0-G9Q |
Rights: | An error occurred on the license name. |
URI: | http://hdl.handle.net/11375/29390 |
Appears in Collections: | Student Publications (Not Graduate Theses) |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
Matira_Kenneth_2023December_MEng.pdf | 4.3 MB | Adobe PDF | View/Open |
Items in MacSphere are protected by copyright, with all rights reserved, unless otherwise indicated.