Skip to content

Commit

Permalink
project page upload
Browse files Browse the repository at this point in the history
  • Loading branch information
@francescodisario
francescodisario committed Oct 4, 2024
1 parent 27200c0 commit f1534cd
Show file tree
Hide file tree
Showing 76 changed files with 11,449 additions and 63 deletions.
138 changes: 101 additions & 37 deletions docs/index.html
View file
Original file line number Diff line number Diff line change
Expand Up @@ -7,6 +7,8 @@
<meta name="keywords" content="Nerfies, D-NeRF, NeRF">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>Boost Your NeRF </title>
<!-- Bootstrap -->
<link rel="stylesheet" href="./static/css/bootstrap-4.4.1.css">

<!-- Global site tag (gtag.js) - Google Analytics -->
<script async src="https://www.googletagmanager.com/gtag/js?id=G-PYVRSFMDRL"></script>
Expand All @@ -22,24 +24,24 @@
gtag('config', 'G-PYVRSFMDRL');
</script>

<link href="https://fonts.googleapis.com/css?family=Google+Sans|Noto+Sans|Castoro"
rel="stylesheet">

<link href="https://fonts.googleapis.com/css?family=Google+Sans|Noto+Sans|Castoro" rel="stylesheet">
<link rel="stylesheet" href="./static/css/bulma.min.css">
<link rel="stylesheet" href="./static/css/bulma-carousel.min.css">
<link rel="stylesheet" href="./static/css/bulma-slider.min.css">
<link rel="stylesheet" href="./static/css/fontawesome.all.min.css">
<link rel="stylesheet"
href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css">
<link rel="stylesheet" href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css">
<link rel="stylesheet" href="./static/css/index.css">
<link rel="icon" href="./static/images/favicon.svg">
<link rel="icon" href="./static/images/icon.png">

<script src="https://ajax.googleapis.com/ajax/libs/jquery/3.5.1/jquery.min.js"></script>
<script defer src="./static/js/fontawesome.all.min.js"></script>
<script src="./static/js/bulma-carousel.min.js"></script>
<script src="./static/js/bulma-slider.min.js"></script>
<script src="./static/js/index.js"></script>
</head>

<link rel="stylesheet" href="./static/css/dics.original.css">
<script src="./static/js/event_handler.js"></script>
<script src="./static/js/dics.original.js"></script>
<script src="./static/js/video-comparison.js"></script>
<body>


Expand Down Expand Up @@ -81,15 +83,15 @@ <h1 class="title is-1 publication-title">Boost Your NeRF: A Model-Agnostic Mixtu
<span>Paper</span>
</a>
</span>
<span class="link-block">
<!-- <span class="link-block">
<a href="https://www.arxiv.org/abs/2407.10389"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="ai ai-arxiv"></i>
</span>
<span>arXiv</span>
</a>
</span>
</span> -->
<!-- Video Link. -->

<!-- Code Link. -->
Expand All @@ -115,8 +117,11 @@ <h1 class="title is-1 publication-title">Boost Your NeRF: A Model-Agnostic Mixtu

<section class="section">
<div class="container is-max-desktop">

<img src="./static/images/trailer.gif" style="margin-bottom: 80px;">
<!-- Abstract. -->
<div class="columns is-centered has-text-centered">

<div class="column is-four-fifths">
<h2 class="title is-3">Abstract</h2>
<div class="content has-text-justified">
Expand Down Expand Up @@ -144,19 +149,17 @@ <h2 class="title is-3">Abstract</h2>
<section class="hero teaser">
<div class="container is-max-desktop">
<div class="hero-body">
<!-- <video id="teaser" autoplay muted loop playsinline height="100%">
<source src="./static/videos/teaser.mp4"
type="video/mp4">
</video> -->


<img src="static/images/architettura.png">
<h2 class="subtitle has-text-centered">
<br>
<span class="dnerf">Boost Your NeRF</span> framework. A density field is used to compute density values for sampled points along
a ray. A filtering step discards low-density points, routed through gating network G
for expert assignment. Top-K experts compute radiance and density, aggregating and
weighting these values by the corresponding gating probability to get the final values
cf and σf . Volume rendering equation yields pixel colors, and joint optimization refines
our resolution-weighted auxiliary loss, allowing for high-quality and efficient rendering.
<span class="dnerf">Boost Your NeRF</span> framework. We first train <i>M</i> models at different resolutions. From them, we distill a small density field,
which is used to compute density values for sampled points along a ray and to discard points in areas with negligible density. For each of these points,
a gating function computes a probability score, indicating the likelihood of routing the point to each expert. We route each point only to the Top-K (with K equals to 1 or 2)
experts, which compute radiance and density. We then aggregate and weight these values by the corresponding gating probability to obtain the final color and density of the point.
Please note how our framework is model-agnostic, treating models as black boxes (we deliver them input, i.e., a point in space, and obtain density and color).
The volume rendering equation yields pixel colors, and joint optimization refines our resolution-weighted auxiliary loss, enabling high-quality and efficient rendering.
</h2>
</div>
</div>
Expand Down Expand Up @@ -204,15 +207,16 @@ <h3 class="title is-4"></h3>
</div>
<br/>

<p> Typical <i>naive</i> approaches to enhance the reconstruction quality of these
<p> Typical <i>naive</i> approaches to enhance the reconstruction quality of Fast-NeRFs
models include:</p>

<li>Increasing the parameters and resolution of the used data structures (e.g.,
voxel grid, hash grid, etc.).</li>
<li>Increasing the number of parameters in the neural network or the order of
spherical harmonics</li>
voxel grid, hash grid, etc.).
<li>Increasing the number of parameters in the neural network or the order of SHs
<li>Increasing the number of sampled points per ray.</li>




<br>
<p>However the (marginal) increase in reconstruction quality results in a significant increase in computational costs.
Expand All @@ -235,10 +239,12 @@ <h3 class="title is-4"></h3>
<h3 class="title is-4">Results</h3>
<div class="content has-text-justified">
<p>
Our methods can overcome these limitations, while being model-agnostic by design.
Our methods can overcome these limitations, while being model-agnostic by design. Although Top-1 already achieves state-of-the-art performance,
Top-2 strikes an optimal balance between the efficiency of the Top-1 approach and the quality of the Ensemble method, where all experts are used for every input.
</p>
</div>
<div class="content has-text-centered">
<img src="static/images/quantitative-results.png" style="margin-bottom: 50px;">
<img src="static/images/results.png">
</div>

Expand All @@ -248,12 +254,52 @@ <h3 class="title is-4">Results</h3>
<h3 class="title is-4">Scene Decomposition</h3>
<div class="content has-text-justified">
<p>
Our method allows for a frequency-based scene decomposition, where high-res models render complex parts of the scene; whereas low-res models render low-frequency parts.
Our method allows for a frequency-based scene decomposition, where high-res models render complex parts of the scene; whereas low-res models render low-frequency parts. The output of the gate
module for each model is visualized on the left; on the right, the rendered part of the scene for each model is shown. Outputs are arranged in increasing order of resolution.
</p>
</div>
<div class="content has-text-centered">
<img src="static/images/partials.png">
</div>

<div class="container">
<div class="content has-text-centered">
<ul class="nav nav-tabs nav-fill nav-justified" id="object-scale-recon">
<li class="nav-item">
<a class="nav-link active" onclick="objectSceneEvent(0)">Lego</a>
</li>
<li class="nav-item">
<a class="nav-link" onclick="objectSceneEvent(1)">Mic</a>
</li>
<li class="nav-item">
<a class="nav-link" onclick="objectSceneEvent(2)">Ship</a>
</li>
<li class="nav-item">
<a class="nav-link" onclick="objectSceneEvent(3)">Hotdog</a>
</li>
<li class="nav-item">
<a class="nav-link" onclick="objectSceneEvent(4)">Ficus</a>
</li>
<li class="nav-item">
<a class="nav-link" onclick="objectSceneEvent(5)">Materials</a>
</li>
<li class="nav-item">
<a class="nav-link" onclick="objectSceneEvent(6)">Chair</a>
</li>
<li class="nav-item">
<a class="nav-link" onclick="objectSceneEvent(7)">Drums</a>
</li>
</ul>
<div class="b-dics" style="font-weight: 600;">
<img src="./static/images/lego/output_0.gif" alt="Low Res">
<img src="./static/images/lego/output_1.gif" alt="Middle Res">
<img src="./static/images/lego/output_2.gif" alt="High Res">

</div>
</div>
</div>


</div>

</div>
Expand All @@ -272,15 +318,41 @@ <h3 class="title is-4">High-Quality Rendering</h3>
<div class="content has-text-centered">
<img src="static/images/quality.png">
</div>
</div>





</div>
</div>
</section>



<section class="section">
<div class="container is-max-desktop">

<h2 class="title is-3 has-text-centered">Ablations</h2>
<div class="columns is-centered">

<div class="column is-full-width">

<strong>Gate Resolution and Gate Type</strong>
<p>Our novel gate design is lightweight (comparable to a linear gate) but achieves way superior quality. Furthermore, it achieves good results with a very-low resolution grid.
</p>
<img src="./static/images/abl1.png" style="padding: 25px;">
</div>

<div class="column is-full-width">
<strong>Why Top-2?</strong>
Top-2 strikes an good trade-off between quality and computational costs.
<img src="./static/images/abl2.png">
</div>


</div>
</div>
</section>

<!-- <section class="section" id="BibTeX">
<div class="container is-max-desktop content">
Expand All @@ -295,23 +367,15 @@ <h2 class="title">BibTeX</h2>
</section> -->



<footer class="footer">
<div class="container">

<div class="columns is-centered">
<div class="column is-8">
<div class="content">
<p>
This website is licensed under a <a rel="license"
href="http://creativecommons.org/licenses/by-sa/4.0/">Creative
Commons Attribution-ShareAlike 4.0 International License</a>.
We thank the authors of <a href="https://nerfies.github.io/">Nerfies</a> that kindly open sourced the template of this website.
</p>
<p>
This means you are free to borrow the <a
href="https://github.com/nerfies/nerfies.github.io">source code</a> of this website,
we just ask that you link back to this page in the footer.
Please remember to remove the analytics code included in the header of the website which
you do not want on your website.
</p>
</div>
</div>
Expand Down
Loading

0 comments on commit f1534cd

Please sign in to comment.