1<!DOCTYPE html> 2<html> 3<head> 4 <meta charset="utf-8"> 5 <meta name="description" content="Certified Human Trajectory Prediction"> 6 <meta name="viewport" content="width=device-width, initial-scale=1"> 7 <title>S-Attack</title> 8 9 <!-- Global site tag (gtag.js) - Google Analytics -->
vendor: 66 bytes, lines 9-10
9 10 <script async src="https://www.googletagmanager.com/gtag/js?id=
10G-PYVRSFMDRL
vendor: 14 bytes, lines 10-11
10"></script> 11
11<link href="https://fonts.googleapis.com/css?family=Google+Sans|Noto+Sans|Castoro" 12 rel="stylesheet"> 13 14 <link rel="stylesheet" href="./static/css/bulma.min.css"> 15 <link rel="stylesheet" href="./static/css/bulma-carousel.min.css"> 16 <link rel="stylesheet" href="./static/css/bulma-slider.min.css"> 17 <link rel="stylesheet" href="./static/css/fontawesome.all.min.css"> 18 <link rel="stylesheet" href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css"> 19 <link rel="stylesheet" href="./static/css/index.css"> 20 21
21<script src="https://ajax.googleapis.com/ajax/libs/jquery/3.5.1/jquery.min.js"></script>
21 22
22<script defer src="./static/js/fontawesome.all.min.js"></script>
22 23
23<script src="./static/js/bulma-carousel.min.js"></script>
23 24
24<script src="./static/js/bulma-slider.min.js"></script>
24 25
25<script src="./static/js/index.js"></script>
25 26</head> 27<body> 28 29<nav class="navbar" role="navigation" aria-label="main navigation"> 30 <div class="navbar-brand"> 31 <a role="button" class="navbar-burger" aria-label="menu" aria-expanded="false"> 32 <span aria-hidden="true"></span> 33 <span aria-hidden="true"></span> 34 <span aria-hidden="true"></span> 35 </a> 36 </div> 37</nav> 38 39 40<section class="hero"> 41 <div class="hero-body"> 42 <div class="container"> 43 <div class="columns is-centered"> 44 <div class="column has-text-centered"> 45 <h1 class="title is-1 publication-title">Certified Human Trajectory Prediction</h1> 46 <div class="is-size-5 publication-authors"> 47 <span class="author-block"> 48 <a href="http://mohammadhossein-bahari.github.io">Mohammadhossein Bahari*</a></sup>,</span> 49 <span class="author-block"> 50 <a href="https://saeedsaadatnejad.github.io">Saeed Saadatnejad*</a></sup>,</span> 51 <span class="author-block"> 52 <a href="">Amirhossein Askari Farsangi</a></sup>,</span> 53 </div> 54 <div class="is-size-5 publication-authors"> 55 <a href="http://smoosavi.me/"> Seyed-Mohsen Moosavi-Dezfooli</a></sup>,</span> 56 <span class="author-block"> 57 <a href="https://people.epfl.ch/alexandre.alahi?lang=en">Alexandre Alahi</a></sup>, 58 </span> 59 </div> 60 61 <div class="is-size-5 publication-authors"> 62 <span class="author-block"></sup>IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025</span> 63 </div> 64 65 <div class="column has-text-centered"> 66 <div class="publication-links"> 67 <!-- arXiv Link. --> 68 <span class="link-block"> 69 <a href="http://arxiv.org/abs/2403.13778" 70 class="external-link button is-normal is-rounded is-dark"> 71 <span class="icon"> 72 <i class="ai ai-arxiv"></i> 73 </span> 74 <span>Paper</span> 75 </a> 76 </span> 77 <!-- Code Link. --> 78 <span class="link-block"> 79 <a href="https://github.com/vita-epfl/s-attack/tree/master/certified" 80 class="external-link button is-normal is-rounded is-dark"> 81 <span class="icon"> 82 <i class="fab fa-github"></i> 83 </span> 84 <span>Code</span> 85 </a> 86 </span> 87 </div> 88 89 </div> 90 </div> 91 </div> 92 </div> 93 </div> 94</section> 95 96 97 98<section class="section"> 99 <div class="container"> 100 <div class="columns is-centered has-text-centered"> 101 <div class="column is-two-thirds"> 102 <div style="align: left; text-align:center;"> 103 <img src="./static/figures/pull.png" width="500"/> 104 <div class="caption"> 105 The top part showcases the outputs of a standard trajectory prediction model, revealing unbounded predictions with noisy inputs. 106 In contrast, the bottom part demonstrates the outputs of our trajectory predictor with guaranteed robustness. 107 The model provides certified bounds (blue boxes) on the predicted outputs, ensuring that outputs remain within guaranteed regions despite input noise. 108 </div> 109 </div> 110 </div> 111 </div> 112 </div> 113</section> 114 115<section class="section"> 116 <div class="container"> 117 <div class="columns is-centered has-text-centered"> 118 <div class="column is-two-thirds"> 119 <div style="align: left; text-align:center;"> 120 <video width="100%" controls> 121 <source src="./static/figures/certified_web.mp4" type="video/mp4"> 122 Your browser does not support the video tag! 123 </video> 124 </div> 125 </div> 126 </div> 127 </div> 128</section> 129 130<section class="section"> 131 <div class="container"> 132 133 <!-- Abstract. --> 134 <div class="columns is-centered has-text-centered"> 135 <div class="column is-two-thirds"> 136 <h2 class="title is-2">Abstract</h2> 137 <div class="content has-text-justified"> 138 <p> 139 Predicting human trajectories is essential for the safe operation of autonomous vehicles, yet current data-driven models often lack robustness in case of noisy inputs such as adversarial examples or imperfect observations. 140 Although some trajectory prediction methods have been developed to provide empirical robustness, these methods are heuristic and do not offer guaranteed robustness. 141 In this work, we propose a certification approach tailored for trajectory prediction that provides guaranteed robustness. 142 To this end, we address the unique challenges associated with trajectory prediction, such as unbounded outputs and multi-modality. 143 To mitigate the inherent performance drop through certification, we propose a diffusion-based trajectory denoiser and integrate it into our method. 144 Moreover, we introduce new certified performance metrics to reliably measure the trajectory prediction performance. 145 Through comprehensive experiments, we demonstrate the accuracy and robustness of the certified predictors and highlight their advantages over the non-certified ones. 146 </div> 147 </div> 148 </div> 149 </div> 150 <!--/ Abstract. --> 151</section> 152 153<section class="section" id="BibTeX">
154 <div class="container content"> 155 <h2 class="title">BibTeX</h2> 156 <pre><code>@InProceedings{bahari2025certified, 157 author = {Bahari, Mohammadhossein and Saadatnejad, Saeed and Askari Farsangi, Amirhossein and Moosavi-Dezfooli, Seyed-Mohsen and Alahi, Alexandre}, 158 title = {Certified Human Trajectory Prediction}, 159 booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, 160 year = {2025}, 161} 162 </code></pre> 163 </div> 164</section> 165 166<footer class="footer"> 167 <div class="columns is-centered has-text-centered"> 168 <div class="is-size-3 publication-authors"> 169 <span class="author-block"></sup>Stay tunned for more updates! 170 <br>Feel free to contact us: {saeed.saadatnejad, mohammadhossein.bahari} at epfl.ch 171 </span> 172 </div> 173 </div> 174</footer> 175 176<footer class="footer"> 177 <div class="columns is-centered has-text-centered"> 178 <div class="column is-two-thirds"> 179 <div class="caption"> This page is created using this <a href="https://github.com/nerfies/nerfies.github.io">source code</a>. 180 </div> 181 </div> 182 </div> 183</footer> 184 185</body> 186</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.