<!DOCTYPE html>
<html lang='en'>
<head>
<meta charset='utf-8'>
<meta name='viewport' content='width=device-width, initial-scale=1, shrink-to-fit=no'>
<meta name='description' content='Sylvain Calinon'>
<meta name='author' content='Sylvain Calinon'>
<meta name='keywords' content='Sylvain Calinon, robotics, machine learning, robot learning, human-robot interaction, human-robot collaboration, learning from demonstration, programming by demonstration, artificial intelligence, Idiap, model-based optimization, Riemannian geometry, tensor methods, tensor factorization, optimal control, ergodic control'>
<meta name='theme-color' content='#343a40'>
<link rel='icon' href='images/favicon.ico'>
<title>Sylvain Calinon</title>
<link href='css/bootstrap.min.css' rel='stylesheet'>
<link href='css/main-template.css' rel='stylesheet'>
<link href='font-awesome/css/font-awesome.min.css' rel='stylesheet'>
<link href='https://fonts.googleapis.com/css?family=Lobster|Raleway' rel='stylesheet'>
<link rel='stylesheet' href='https://cdn.jsdelivr.net/npm/katex@0.16.2/dist/katex.min.css' integrity='sha384-bYdxxUwYipFNohQlHt0bjN/LCpueqWz13HufFEV1SUatKs1cm4L6fFgCi1jT643X' crossorigin='anonymous'>
<script defer src='https://cdn.jsdelivr.net/npm/katex@0.16.2/dist/katex.min.js' integrity='sha384-Qsn9KnoKISj6dI8g7p1HBlNpVx0I8p1SvlwOldgi3IorMle61nQy4zEahWYtljaz' crossorigin='anonymous'></script>
<script defer src='https://cdn.jsdelivr.net/npm/katex@0.16.2/dist/contrib/auto-render.min.js' integrity='sha384-+VBxd3r6XgURycqtZ117nYw44OOcIax56Z4dCRWbxyPt0Koah1uHoK0o4+/RRE05' crossorigin='anonymous'></script>
<script>
let macros = {
'\\tp': '\\text{\\tiny{#1}}',
'\\trsp' : '\\top',
'\\psin' : '\\dagger',
'\\eqref': '\\href{###1}{(\\text{#1})}',
'\\ref': '\\href{###1}{\\text{#1}}',
'\\label': '\\htmlId{#1}{}'
};
document.addEventListener('DOMContentLoaded', function() {
renderMathInElement(document.body, {
// customised options
trust: (context) => ['\\htmlId', '\\href'].includes(context.command),
macros: macros,
// • auto-render specific keys, e.g.:
delimiters: [
{left: '$$', right: '$$', display: true},
{left: '$', right: '$', display: false},
{left: '\\(', right: '\\)', display: false},
{left: '\\begin{equation}', right: '\\end{equation}', display: true},
{left: '\\begin{equation*}', right: '\\end{equation*}', display: true},
{left: '\\begin{align}', right: '\\end{align}', display: true},
{left: '\\begin{align*}', right: '\\end{align*}', display: true},
{left: '\\begin{alignat}', right: '\\end{alignat}', display: true},
{left: '\\begin{gather}', right: '\\end{gather}', display: true},
{left: '\\begin{CD}', right: '\\end{CD}', display: true},
{left: '\\[', right: '\\]', display: true}
],
// • rendering keys, e.g.:
throwOnError : false
});
});
</script>
</head>
<body>
<nav class='navbar navbar-expand-lg navbar-dark bg-dark fixed-top'>
<button class='navbar-toggler navbar-toggler-right' type='button' data-toggle='collapse' data-target='#navbarsExampleDefault' aria-controls='navbarsExampleDefault' aria-expanded='false' aria-label='Toggle navigation'>
<span class='navbar-toggler-icon'></span>
</button>
<a class='navbar-brand' href='index.htm'>
<h3 style='display: inline-block;'>Sylvain Calinon</h3>
<img src='images/pbd-thumbnail02.png' style='display: inline-block; margin-left: 10px;'>
</a>
<div class='collapse navbar-collapse' id='navbarsExampleDefault'>
<ul class='navbar-nav mr-auto'>
<li class='nav-item'><a class='nav-link' href='index.htm'>News</a></li>
<li class='nav-item'><a class='nav-link' href='cv.htm'>CV</a></li>
<li class='nav-item'><a class='nav-link' href='research.htm'>Research</a></li>
<li class='nav-item'><a class='nav-link' href='publications.htm'>Publications</a></li>
<li class='nav-item'><a class='nav-link' href='teaching.htm'>Teaching</a></li>
<li class='nav-item'><a class='nav-link' href='book.htm'>Book</a></li>
<li class='nav-item'><a class='nav-link' href='videos.htm'>Videos</a></li>
<li class='nav-item'><a class='nav-link' href='codes.htm'>Codes</a></li>
<li class='nav-item'><a class='nav-link' href='open-positions.htm'>Open positions </a></li>
<li class='nav-item'><a class='nav-link' href='contact.htm'>Contact/Links</a></li>
</ul>
</div>
</nav>
<div class='container'>
<div class='main-template'>
<div class='row'>
<div class='col-md-2 text-center'><img class='borderthumb' src='images/3104.jpg'></div>
<div class='col-md-10'>
<h5>Xue, T., Razmjoo, A., Shetty, S. and Calinon, S. (2024)</h5>
<h5><strong>Robust Manipulation Primitive Learning via Domain Contraction</strong></h5>
<h5>In Proc. Conference on Robot Learning (CoRL).</h5>
<div class='row'>
<div class='col-md-4'>
<a href='papers/Xue-CORL2024.pdf' target='_BLANK'><i class='fa fa-file-pdf-o fa-fw'></i> PDF</a>
</div>
<div class='col-md-4'>
<a href='papers/Xue-CORL2024.bib' class='bibtex' target='_BLANK'><i class='fa fa-file-text-o fa-fw'></i> Reference as bib file</a>
</div>
</div></div></div>
<h3>Abstract</h3>
<p>Contact-rich manipulation plays an important role in everyday life, but uncertain parameters pose significant challenges to model-based planning and control. To address this issue, domain adaptation and domain randomization have been proposed to learn robust policies. However, they either lose the generalization ability to diverse instances or perform conservatively due to neglecting instance-specific information. In this paper, we propose a bi-level approach to learn robust manipulation primitives, including parameter-augmented policy learning using multiple models with tensor approximation, and parameter-conditioned policy retrieval through domain contraction. This approach unifies domain randomization and domain adaptation, providing optimal behaviors while keeping generalization ability. We validate the proposed method on three contact-rich manipulation primitives: hitting, pushing, and reorientation. The experimental results showcase the superior performance of our approach in generating robust policies for instances with diverse physical parameters.</p>
<h3>Bibtex reference</h3>
<pre>@inproceedings{Xue24CORL,
author={Xue, T. and Razmjoo, A. and Shetty, S. and Calinon, S.},
title={Robust Manipulation Primitive Learning via Domain Contraction},
booktitle={Proc.\ Conference on Robot Learning ({CoRL})},
year={2024}
}
</pre>
<center><a class='btn btn-primary' href='publications.htm'>Go back to the list of publications</a></center></div><!-- /.main-template -->
</div><!-- /.container -->
<!-- Bootstrap core JavaScript placed at the end of the document so the pages load faster -->
<script src="https://code.jquery.com/jquery-3.1.1.slim.min.js" integrity="sha384-A7FZj7v+d/sdmMqp/nOQwliLvUsJfDHW+k9Omg/a/EheAdgtzNs3hpfag6Ed950n" crossorigin="anonymous"></script>
<script>window.jQuery || document.write('<script src="../../assets/js/vendor/jquery.min.js"><\/script>')</script>
<script src="https://cdnjs.cloudflare.com/ajax/libs/tether/1.4.0/js/tether.min.js" integrity="sha384-DztdAPBWPRXSA/3eYEEUWrWCy7G5KFbe8fFjk5JAIxUYHKkDx6Qin1DkWx51bBrb" crossorigin="anonymous"></script>
<script src="js/bootstrap.min.js"></script>
<!-- IE10 viewport hack for Surface/desktop Windows 8 bug -->
<!--<script src="assets/js/ie10-viewport-bug-workaround.js"></script>-->
</body>
</html>