Skip to content

Commit

Permalink
Add robot localization experiment, add code link and citation
Browse files Browse the repository at this point in the history
  • Loading branch information
janapavlasek committed Mar 11, 2024
1 parent 417fe81 commit 1772471
Show file tree
Hide file tree
Showing 3 changed files with 25 additions and 24 deletions.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
49 changes: 25 additions & 24 deletions projects/stein-bp/index.html
Original file line number Diff line number Diff line change
Expand Up @@ -93,11 +93,11 @@ <h3><a href="https://fabioramos.github.io/">Fabio Ramos</a></h3>
<p>
<a href="https://arxiv.org/abs/2311.16916"><i class="far fa-file-pdf"></i>&nbsp;&nbsp;Preprint</a>&nbsp;&nbsp;
<a href="https://youtu.be/buRdZnahwBc"><i class="fab fa-youtube"></i>&nbsp;&nbsp;Video</a>&nbsp;&nbsp;
<a href="https://openreview.net/pdf?id=fDsqkXx1Ild"><i class="far fa-file-pdf"></i>&nbsp;&nbsp;Workshop Paper</a>&nbsp;&nbsp;
<!-- <a href="https://openreview.net/pdf?id=fDsqkXx1Ild"><i class="far fa-file-pdf"></i>&nbsp;&nbsp;Workshop Paper</a>&nbsp;&nbsp; -->
<!-- Optional. -->
<!-- <a href="#"><i class="fas fa-chalkboard-teacher"></i>&nbsp;&nbsp;Talk</a>&nbsp;&nbsp; -->
<!-- <a href="#"><i class="fas fa-code"></i>&nbsp;&nbsp;Code</a>&nbsp;&nbsp; -->
<!-- <a href="#cite"><i class="fas fa-quote-right"></i>&nbsp;&nbsp;Cite</a> -->
<a href="https://github.com/janapavlasek/multi-robot-svbp"><i class="fab fa-github"></i>&nbsp;&nbsp;Code</a>&nbsp;&nbsp;
<a href="#cite"><i class="fas fa-quote-right"></i>&nbsp;&nbsp;Cite</a>
</p>
</div>

Expand All @@ -108,19 +108,19 @@ <h3><a href="https://fabioramos.github.io/">Fabio Ramos</a></h3>
<h2>Multi-Robot Perception</h2>
</header>

<p style="margin-bottom: 1em;">We implement our SVBP on a graph problem simulating perception under challenging noise conditions. The goal is to localize each node in the "spider" where the observation is the multi-modal Gaussian shown in the video.</p>
<p style="margin-bottom: 1em;">We implement our SVBP on a graph problem simulating multi-robot perception under challenging noise conditions. The goal is to localize each robot given the displacement between neighboring robots and a the multi-modal Gaussian observation shown in the video.</p>

<div class="row aln-center exp-grid">
<div class="col-5 col-5-large col-5-medium col-8-small">
<div class="col-5 col-5-large col-6-medium col-12-small">
<p style="text-align: center; margin: 0;"><b>SVBP (Ours)</b></p>
<div class="image experiment centered">
<img src="images/exp/spider/sbp_50_spider_30_1.gif" alt="" />
<img src="images/exp/multi-loc/sbp_50_robots_30_3.gif" alt="" />
</div>
</div>
<div class="col-5 col-5-large col-5-medium col-8-small">
<div class="col-5 col-5-large col-6-medium col-12-small">
<p style="text-align: center; margin: 0;"><b>Particle Belief Propagation</b></p>
<div class="image experiment centered">
<img src="images/exp/spider/pbp_nodes_50_spider_30_1.gif" alt="" />
<img src="images/exp/multi-loc/pbp_nodes_50_robots_30_3.gif" alt="" />
</div>
</div>
</div>
Expand All @@ -137,39 +137,39 @@ <h2>Multi-Robot Planning</h2>
<p>We apply SVBP to the multi-robot planning problem. Below are video results for each scene in the dataset.</p>

<div class="row aln-center exp-grid">
<div class="col-4 col-4-large col-5-medium col-8-small">
<div class="col-4 col-4-large col-6-medium col-12-small">
<div class="image experiment centered">
<video controls>
<source src="images/exp/planar/svbp/svbp_swap_rooms_0.mp4" type="video/mp4">
Your browser does not support the video tag.
</video>
</div>
</div>
<div class="col-4 col-4-large col-5-medium col-8-small">
<div class="col-4 col-4-large col-6-medium col-12-small">
<div class="image experiment centered">
<video controls>
<source src="images/exp/planar/svbp/svbp_obstacles_mix_circle_0.mp4" type="video/mp4">
Your browser does not support the video tag.
</video>
</div>
</div>
<div class="col-4 col-4-large col-5-medium col-8-small">
<div class="col-4 col-4-large col-6-medium col-12-small">
<div class="image experiment centered">
<video controls>
<source src="images/exp/planar/svbp/svbp_squares_3x3_cross_0.mp4" type="video/mp4">
Your browser does not support the video tag.
</video>
</div>
</div>
<div class="col-4 col-4-large col-5-medium col-8-small">
<div class="col-4 col-4-large col-6-medium col-12-small">
<div class="image experiment centered">
<video controls>
<source src="images/exp/planar/svbp/svbp_squares_swap_0.mp4" type="video/mp4">
Your browser does not support the video tag.
</video>
</div>
</div>
<div class="col-4 col-4-large col-5-medium col-8-small">
<div class="col-4 col-4-large col-6-medium col-12-small">
<div class="image experiment centered">
<video controls>
<source src="images/exp/planar/svbp/svbp_end_of_lessons_0.mp4" type="video/mp4">
Expand All @@ -184,7 +184,7 @@ <h3>Baselines</h3>
<p>We implement different baselines to compare our algorithm. Click through the examples for each below.</p>

<div class="row aln-center">
<div class="col-5 col-5-large col-5-medium col-8-small project-author">
<div class="col-5 col-5-large col-6-medium col-12-small project-author">
<div class="slideshow-wrapper">
<div class="slideshow-container" id="gabp-mpc-slideshow">
<div class="image slide">
Expand Down Expand Up @@ -225,7 +225,7 @@ <h3>Baselines</h3>
<h3>Gaussian Belief Propagation</h3>
<p>A version of our algorithm with a Gaussian trajectory distribution instead of a nonparametric one. The GaBP baseline is more likely to fall into local minima.</p>
</div>
<div class="col-5 col-5-large col-5-medium col-8-small project-author">
<div class="col-5 col-5-large col-6-medium col-12-small project-author">
<div class="slideshow-wrapper">
<div class="slideshow-container" id="orca2-mpc-slideshow">
<div class="image slide">
Expand Down Expand Up @@ -266,7 +266,7 @@ <h3>Gaussian Belief Propagation</h3>
<h3>ORCA (20 cm radius)</h3>
<p>The popular ORCA method (van den Berg 2011) is efficient but not as smooth as SVBP and prone to deadlock.</p>
</div>
<div class="col-5 col-5-large col-5-medium col-8-small project-author">
<div class="col-5 col-5-large col-6-medium col-12-small project-author">
<div class="slideshow-wrapper">
<div class="slideshow-container" id="orca4-mpc-slideshow">
<div class="image slide">
Expand Down Expand Up @@ -310,20 +310,21 @@ <h3>ORCA (40 cm radius)</h3>
</div>
</article>

<!-- <article class="box post">
<article class="box post">
<header id="cite">
<h2>Citation</h2>
</header>

The whitespace between <pre> tags will be preserved exactly how it is in the code.
<!-- The whitespace between <pre> tags will be preserved exactly how it is in the code. -->
<div class="citation">
<pre><code>@inproceedings{pavlasek2023svbp,
author = {Last, First and Last, First and ...},
title = {Paper title},
year = {2020}
<pre><code>@inproceedings{pavlasek2024stein,
title={Stein Variational Belief Propagation for Multi-Robot Coordination},
author={Pavlasek, Jana and Mah, Joshua Jing Zhi and Xu, Ruihan and Jenkins, Odest Chadwicke and Ramos, Fabio},
booktitle={Robotics and Automation Letters (RA-L)},
year={2024}
}</code></pre>
</div>
</article> -->
</article>

</div>
</section>
Expand All @@ -333,7 +334,7 @@ <h2>Citation</h2>
<!-- Copyright -->
<div id="copyright">
<ul class="links">
<li>&copy; 2023. All rights reserved.</li><li>Design: Dopetrope by
<li>&copy; 2024. All rights reserved.</li><li>Design: Dopetrope by
<a href="http://twitter.com/ajlkn">AJ</a> for <a href="http://html5up.net/">HTML5 UP</li>
</ul>
</div>
Expand Down

0 comments on commit 1772471

Please sign in to comment.