1<!DOCTYPE html> 2<html> 3 4 <head> 5 <meta charset="utf-8"> 6 <meta http-equiv="X-UA-Compatible" content="IE=edge"> 7 <meta name="viewport" content="width=device-width, initial-scale=1"> 8 <title>EAIS Lab - Publications</title> 9 <meta name="description" content="EAIS Lab -- Publications."> 10 <link rel="stylesheet" href="/css/main.css"> 11 <link rel="canonical" href="/publications/"> 12 13 14<!--
14<script>
vendor: 345 bytes, lines 14-20
14 15 (function(i,s,o,g,r,a,m){i['GoogleAnalyticsObject']=r;i[r]=i[r]||function(){ 16 (i[r].q=i[r].q||[]).push(arguments)},i[r].l=1*new Date();a=s.createElement(o), 17 m=s.getElementsByTagName(o)[0];a.async=1;a.src=g;m.parentNode.insertBefore(a,m) 18 })(window,document,'script','https://www.google-analytics.com/analytics.js','ga'); 19 20 ga('create', '
20UA-82472331-1
vendor: 39 bytes, lines 20-22
20', 'auto'); 21 ga('send', 'pageview'); 22
23</script>
23 --> 24<!-- Google tag (gtag.js) -->
vendor: 64 bytes, lines 24-25
24 25<script async src="https://www.googletagmanager.com/gtag/js?id=
25G-6324B4RBEB
vendor: 12 bytes, line 25
25"></script>
26<script> 27
vendor: 134 bytes, lines 27-31
27window.dataLayer = window.dataLayer || []; 28 function gtag(){dataLayer.push(arguments);} 29 gtag('js', new Date()); 30 31 gtag('config', '
31G-6324B4RBEB
vendor: 4 bytes, line 31
31');
32</script>
32 33 34 35 36</head> 37 38 39 <body> 40 41 <div class="navbar navbar-default navbar-fixed-top" role="navigation"> 42 <div class="container-fluid"> 43 <div class="navbar-header"> 44 <button type="button" class="navbar-toggle collapsed" data-toggle="collapse" data-target="#navbar-collapse-1" aria-expanded="false"> 45 <span class="sr-only">Toggle navigation</span> 46 <span class="icon-bar"></span> 47 <span class="icon-bar"></span> 48 <span class="icon-bar"></span> 49 </button> 50 51 <a class="navbar-brand" href="/">Embedded AI Systems Lab @ EMU</a> 52 </div> 53 <div class="collapse navbar-collapse" id="navbar-collapse-1"> 54 <ul class="nav navbar-nav navbar-right"> 55 <li><a href="/">Home</a></li> 56 <li><a href="/team">Team</a></li> 57 <li><a href="/vacancies">Vacancies</a></li> 58 <li><a href="/publications">Publications</a></li> 59 <li><a href="/research">Research</a></li> 60<!-- <li><a href="/pictures">(Pics)</a></li> --> 61 </ul> 62 </div> 63 </div> 64</div> 65 66 67 <div class="container-fluid"> 68 <div class="row"> 69 <div id="gridid" class="col-sm-12"> 70 <h1 id="journal-articles">Journal articles</h1> 71 72<div class="row"> 73 74 <div class="col-sm-6 clearfix"> 75 <div class="well"> 76 <pubtit>Planogen: A Procedural Generation Framework for Dynamic VR Research Environments</pubtit> 77 <!-- <img src="/images/pubpic/" class="img-responsive" width="33%" style="float: left" /> --> 78 <p>Introducing a modular procedural generation plug-in for the Unity game engine, for the rapid generation of varied and interactive aircraft cabin environments populated with diverse virtual passengers.</p> 79 <p><em>Kaitlyn Tracy, Lazaros Rafail Kouzelis, Rami Dari, Ourania Spantidi</em></p> 80 <p><strong>Virtual Worlds 2025</strong></p> 81 <p class="text-danger"><strong> </strong></p> 82 <p> </p> 83 </div> 84 </div> 85 86 <div class="col-sm-6 clearfix"> 87 <div class="well"> 88 <pubtit>Impact of GPT-Driven Teaching Assistants in VR Learning Environments</pubtit> 89 <!-- <img src="/images/pubpic/" class="img-responsive" width="33%" style="float: left" /> --> 90 <p>Evaluating the impact of a GPT teaching assistant in a VR classroom.</p> 91 <p><em>Kaitlyn Tracy, Ourania Spantidi</em></p> 92 <p><strong>IEEE Transactions on Learning Technologies 2025</strong></p> 93 <p class="text-danger"><strong> </strong></p> 94 <p> </p> 95 </div> 96 </div> 97 98</div> 99 100<div class="row"> 101 102 <div class="col-sm-6 clearfix"> 103 <div class="well"> 104 <pubtit>Enhancing historical extended reality experiences: Prompt engineering strategies for AI-generated dialogue</pubtit> 105 <!-- <img src="/images/pubpic/" class="img-responsive" width="33%" style="float: left" /> --> 106 <p>Introducing refined prompt engineering techniques to reduce factual inaccuracies and enhance the emotional resonance of AI-generated dialogues.</p> 107 <p><em>Lazaros Rafail Kouzelis, Ourania Spantidi</em></p> 108 <p><strong>Applied Sciences 2024</strong></p> 109 <p class="text-danger"><strong> </strong></p> 110 <p> </p> 111 </div> 112 </div> 113 114</div> 115 116<p> Â </p> 117 118<h1 id="conference-articles">Conference articles</h1> 119 120<div class="row"> 121 122 <div class="col-sm-6 clearfix"> 123 <div class="well"> 124 <pubtit>Backbone Choice Matters: Human and Perceptual Evaluation of CNN-Based Neural Style Transfer</pubtit> 125 <!-- <img src="/images/pubpic/" class="img-responsive" width="33%" style="float: left" /> --> 126 <p>This paper investigates how the choice of CNN backbone affects neural style transfer, finding that stronger backbones generally produce images preferred by users, while standard image-quality metrics do not fully capture perceived artistic quality or style similarity.</p> 127 <p><em>Andie Bradley, Ourania Spantidi</em></p> 128 <p><strong>International Conference on Machine Learning and Applications 2026</strong></p> 129 <p class="text-danger"><strong> </strong></p> 130 <p> </p> 131 </div> 132 </div> 133 134 <div class="col-sm-6 clearfix"> 135 <div class="well"> 136 <pubtit>Detecting Gray Failures in Edge Networks with Lightweight Unsupervised Learning</pubtit> 137 <!-- <img src="/images/pubpic/" class="img-responsive" width="33%" style="float: left" /> --> 138 <p>This paper presents a lightweight unsupervised learning system that detects gray failures in heterogeneous edge networks and triggers proactive failover, reducing HTTP latency by up to 72% in controlled experiments.</p> 139 <p><em>William Britton, Ourania Spantidi</em></p> 140 <p><strong>International Conference on Machine Learning and Applications 2026</strong></p> 141 <p class="text-danger"><strong> </strong></p> 142 <p> </p> 143 </div> 144 </div> 145 146</div> 147 148<div class="row"> 149 150 <div class="col-sm-6 clearfix"> 151 <div class="well"> 152 <pubtit>'Are We Going To Crash?': The Effect of an AI Presence in User Experience during a VR Flight</pubtit> 153 <!-- <img src="/images/pubpic/" class="img-responsive" width="33%" style="float: left" /> --> 154 <p>A pilot study that examines the effect of an AI presence while experiencing turbulence during a VR flight.</p> 155 <p><em>Kaitlyn Tracy, William Dalian, Rami Dari, Ourania Spantidi</em></p> 156 <p><strong>International Symposium on Visual Computing 2025</strong></p> 157 <p class="text-danger"><strong> </strong></p> 158 <p> </p> 159 </div> 160 </div> 161 162 <div class="col-sm-6 clearfix"> 163 <div class="well"> 164 <pubtit>Toward Dynamic NPC Interactions: Integrating GPT-Driven Agents in 3D Virtual Environments</pubtit> 165 <!-- <img src="/images/pubpic/" class="img-responsive" width="33%" style="float: left" /> --> 166 <p>This paper introduces a novel framework that extends the capabilities of GPT to control comprehensive NPC behavior within Unity.</p> 167 <p><em>Jorde Stojanov, Lazaros Rafail Kouzelis, Ourania Spantidi</em></p> 168 <p><strong>International Symposium on Visual Computing 2024</strong></p> 169 <p class="text-danger"><strong> </strong></p> 170 <p> </p> 171 </div> 172 </div> 173 174</div> 175 176<div class="row"> 177 178 <div class="col-sm-6 clearfix"> 179 <div class="well"> 180 <pubtit>Synthesizing play-ready VR scenes with natural language prompts through GPT API</pubtit> 181 <!-- <img src="/images/pubpic/" class="img-responsive" width="33%" style="float: left" /> --> 182 <p>This work utilizes the GPT API to interpret natural language prompts and generate detailed, VR-ready scenes within Unity3D.</p> 183 <p><em>Lazaros Rafail Kouzelis, Ourania Spantidi</em></p> 184 <p><strong>International Symposium on Visual Computing 2023</strong></p> 185 <p class="text-danger"><strong> </strong></p> 186 <p> </p> 187 </div> 188 </div> 189 190</div> 191 192<p> Â </p> 193 194 195</div> 196 197 </div> 198 </div> 199 200 <div id="footer" class="panel"> 201 <div class="panel-footer"> 202 <div class="container-fluid"> 203 <div class="row"> 204 205 <div class="col-sm-4"> 206 <p>© 2024 EAS Lab. Layout forked by <a href="https://www.allanlab.org/">Allan's lab</a>.</p> 207 </div> 208 209 <div class="col-sm-4"> 210 Contact:<br /> 211 Pray Harrold, 5th floor<br /> 212 Ypsilanti, Michigan<br /> 213 USA <br /> 214 </div> 215 216 <div class="col-sm-4 footer-logo"> 217 <img src="/images/EAIS_Lab_Logo-1-removebg-preview.png" alt="EAIS Lab logo"> 218 </div> 219 220 </div> 221 </div> 222 </div> 223</div> 224
225<script src="/js/bootstrap.min.js"></script>
225 226 227 228 </body> 229 230</html>
Line numbers count LF bytes from the start of the resource, as the search results do. Vendor segments are library code the classifier recognised; they are stored but not indexed. Bytes are shown as Latin1 characters, one per byte.