Skip to content

Commit

Permalink
update mobile rendering
Browse files Browse the repository at this point in the history
  • Loading branch information
@hao-pt
hao-pt committed Oct 22, 2024
1 parent 5cf1442 commit 4912066
Show file tree
Hide file tree
Showing 3 changed files with 46 additions and 72 deletions.
113 changes: 41 additions & 72 deletions index.html
View file
Original file line number Diff line number Diff line change
Expand Up @@ -108,7 +108,7 @@
<div class="container is-max-desktop">
<div class="columns is-centered">
<div class="column has-text-centered">
<h1 class="title is-1 publication-title">DiMSUM <img src="./static/images/dimsum_icon.png" class="logo" width=5.5% />: <span style="color:red;">Di</span>ffusion <span style="color:red;">M</span>amba - A <span style="color:red;">S</span>calable and <span style="color:red;">U</span>nified
<h1 class="title is-1 publication-title">DiMSUM <img src="./static/images/dimsum_icon.png" class="logo" width="50px" />: <span style="color:red;">Di</span>ffusion <span style="color:red;">M</span>amba - A <span style="color:red;">S</span>calable and <span style="color:red;">U</span>nified
Spatial-Frequency <span style="color:red;">M</span>ethod for Image Generation</h1>
<div class="is-size-5 publication-authors">
<span class="author-block">
Expand All @@ -124,10 +124,10 @@ <h1 class="title is-1 publication-title">DiMSUM <img src="./static/images/dimsum
<a href="https://viethoang1512.github.io/">Hoang Phan</a><sup>4</sup>&emsp;
</span>
<span class="author-block">
<a href="https://people.cs.rutgers.edu/~dnm/">Dimitris N. Metaxas</a><sup>3</sup>&emsp;
<a href="https://people.cs.rutgers.edu/~dnm/">Dimitris N. Metaxas</a><sup>2</sup>&emsp;
</span>
<span class="author-block">
<a href="https://sites.google.com/site/anhttranusc/">Anh Tran</a><sup>1</sup>
<a href="https://scholar.google.com/citations?user=FYZ5ODQAAAAJ">Anh Tran</a><sup>1</sup>
</span>
</div>

Expand Down Expand Up @@ -524,11 +524,10 @@ <h2 class="title is-3">Unconditional Generation</h2>

<!-- Motivation -->
<h2 class="title is-3">Why is scanning in frequency space helpful?</h2>
<div class="columns is-centered">
<div class="columns is-centered" style="text-align: center;">
<img src="./static/images/wavelet_vs_spatial_window.png" width="60%" class="scanning" />
</div>
<div class="content has-text-justified">

<p>
Previous state-space models, particularly in processing visual data, failed to effectively address the design choice of scanning order due to their exclusive reliance on spatial processing, neglecting crucial long-range relations in the frequency spectrum.
We propose a novel approach that integrates frequency scanning with the conventional spatial scanning mechanism.
Expand Down Expand Up @@ -595,88 +594,58 @@ <h3 class="title is-4">Globally-shared Transformer Block</h3>
<div class="column is-full-width">
<h2 class="title is-3">Results</h2>
</div>
<div class="columns is-centered">
<div class="column">
<div class="content">
<figure>
<img src="./static/images/celeb256.jpg" class="interpolation-image" width=45% style="margin-right: 60px;" />
<img src="./static/images/celeb512.jpg" class="interpolation-image" width=45% />
<figcaption>Figure 1. Unconditional generation on CelebA HQ</figcaption>
</figure>

</div>
</div>
</div>
<br>
<div class="columns is-centered" style="text-align: center;">
<table>
<tr>
<td><img src="./static/images/celeb256.jpg" width="90%" /> </td>
<td> <img src="./static/images/celeb512.jpg" width="90%" /> </td>
</tr>
<caption style="text-align: center; color: black">Figure 1. Unconditional generation on CelebA HQ 256 & 512</caption>
</table>
</div>
<div class="columns is-centered">
<div class="column" style="margin-left: 13%;">
<div class="content">
<figure>
<img src="./static/images/church256.jpg" class="interpolation-image" width=70% />
<figcaption>Figure 2. Unconditional generation on LSUN Church</figcaption>
</figure>
</div>
<!-- <div class="content">
<figure>
<img src="./static/images/imnet.jpg" class="interpolation-image" width=35% />
<figcaption>Figure 3. Class-conditional generation on ImageNet1k 256</figcaption>
</figure>
</div> -->
<div class="columns is-centered" style="text-align: center;">
<div class="column">
<figure >
<img src="./static/images/training_convergence.png" class="interpolation-image" width="70%"/>
<figcaption>Figure 2. Training convergence on CelebA HQ 256.</figcaption>
</figure>
</div>
<div class="column">
<div class="content" style="margin-left: -42%;">
<figure>
<img src="./static/images/imnet.jpg" class="interpolation-image" width=60% />
<figcaption>Figure 3. Class-conditional generation on ImageNet1k 256</figcaption>
</figure>
</div>
<figure>
<img src="./static/images/church256.jpg" class="interpolation-image" width="90%" />
<figcaption>Figure 3. Unconditional generation on LSUN Church</figcaption>
</figure>
</div>
<!-- <div class="column">
<div class="content">
<figure>
<img src="./static/images/training_convergence.png" class="interpolation-image" width=35% />
<figcaption style="margin-left: 400px; margin-right: 400px;">
Figure 4. Training convergence on CelebA HQ 256.
Our method achieves faster training convergence, requiring fewer than half the training epochs compared to other diffusion models, while delivering a more stable training curve.
</figcaption>
</figure>
</div>
</div> -->
</div>
<div class="columns is-centered" style="text-align: center;">
</div>
<div class="columns is-centered">
<div class="column">
<div class="content">
<figure>
<img src="./static/images/training_convergence.png" class="interpolation-image" width=35% />
<figcaption style="margin-left: 400px; margin-right: 400px;">
Figure 4. Training convergence on CelebA HQ 256.
Our method achieves faster training convergence, requiring fewer than half the training epochs compared to other diffusion models, while delivering a more stable training curve.
</figcaption>
<img src="./static/images/imnet.jpg" class="interpolation-image" width=60% />
<figcaption>Figure 4. Class-conditional generation on ImageNet1k 256</figcaption>
</figure>
</div>
</div>
</div>
</section>

<!-- <section class="section" id="bound">
<div class="container is-max-desktop content">
<h2 class="title">Theoretical analysis: Bounding estimation error</h2>
<div class="content has-text-justified">
We have shown that minimizing the FM objective on latent space controls the Wasserstein distance between the
target density \( p_0 \) and the reconstructed density \( \hat{p}_0 \), which coincides with Fréchet inception
distance (FID), a common metric for image generation.
This means that our latent flow matching is guaranteed to control this metric, given reasonable estimation of \(
\hat{v}(\mathbf{z}_t, t) \). Nonetheless, the analysis also suggests that the quality of latent flow matching
depends on the constants that define the expressivity of the decoders and encoders, which has been observed in
prior research on generative modeling in latent space.
<img src="./static/images/bound.jpeg" class="interpolation-image" />


<section class="section" id="speed">
<div class="container is-max-desktop">
<div class="column is-full-width">
<h2 class="title is-3">Speed</h2>
</div>
</div>
</section> -->
<div class="container is-max-desktop content" style="text-align: center;">
<div>
<img src="./static/images/speed.jpg" class="interpolation-image" width="80%"/>
</div>
<div class="content has-text-justified">
The speed gap between our method and DiT widens as the input resolution increases, highlighting the efficiency of our method for high-resolution synthesis.
</div>
</div>
</section>

<!-- <section class="section" id="Related">
<div class="container is-max-desktop content">
Expand Down
5 changes: 5 additions & 0 deletions static/css/index.css
View file
Original file line number Diff line number Diff line change
Expand Up @@ -251,4 +251,9 @@ background-color: #F2E34C;
--mask-image: repeating-linear-gradient(135deg, #000 0px, #000 1px, transparent 1px, transparent 6px);
-webkit-mask-image: var(--mask-image);
mask-image: var(--mask-image);
}

@media (max-width:600px) {
.about-us-text,
.about-us-images { width:100% }
}
Binary file added static/images/speed.jpg
View file
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.

0 comments on commit 4912066

Please sign in to comment.