Skip to content

Commit

Permalink
add images
Browse files Browse the repository at this point in the history
  • Loading branch information
maobenz committed Aug 23, 2024
1 parent 8a8d423 commit 783bea7
Showing 1 changed file with 22 additions and 26 deletions.
48 changes: 22 additions & 26 deletions index.html
Original file line number Diff line number Diff line change
Expand Up @@ -102,7 +102,7 @@
<link href="./assets/cover.css" rel="stylesheet">
<script src="https://ajax.googleapis.com/ajax/libs/jquery/3.7.0/jquery.min.js"></script>
</head>
<body class="text-center text-bg-dark">
<body class="text-center text-bg-dark">
<svg xmlns="http://www.w3.org/2000/svg" class="d-none">
<symbol id="check2" viewBox="0 0 16 16">
<path d="M13.854 3.646a.5.5 0 0 1 0 .708l-7 7a.5.5 0 0 1-.708 0l-3.5-3.5a.5.5 0 1 1 .708-.708L6.5 10.293l6.646-6.647a.5.5 0 0 1 .708 0z"/>
Expand All @@ -121,9 +121,6 @@



<body class="text-center2 text-bg-white">
</body>

<main>

<section class="text-center container min-vh-0">
Expand Down Expand Up @@ -194,29 +191,28 @@ <h1>Abstract</h1>
We present a unified transformer, i.e., Show-o, that unifies multimodal understanding and generation. Unlike fully autoregressive models, Show-o unifies autoregressive and (discrete) diffusion modeling to adaptively handle inputs and outputs of various and mixed modalities. The unified model flexibly supports a wide range of vision-language tasks including visual question-answering, text-to-image generation, text-guided inpainting/extrapolation, and mixed-modality generation. Across various benchmarks, it demonstrates comparable or superior performance to existing individual models with an equivalent or larger number of parameters tailored for understanding or generation. This significantly highlights its potential as a next-generation foundation model.
</p>



</div>
</div>


<div class="album py-5 bg-body-tertiary text-dark ">

<div class="container col-lg-6 mx-auto">
<div class="album py-5 bg-body-tertiary text-dark" >

<div class="container col-lg-6 mx-auto" style="background-color: #f0f0f0;">

<h1>What's new of Show-o?</h1>
<h1 style="background-color: #f0f0f0;">What's new of Show-o?</h1>

<img src="./assets/images/comparison.png" style="width: 100%;">
</div>
</div>



<div class="album py-5 bg-body-tertiary text-dark">
<div class="container col-lg-6 mx-auto">
<div class="album py-5 bg-body-tertiary text-dark " >
<div class="container col-lg-6 mx-auto" style="background-color: #f0f0f0;">

<h1>Method</h1>
<p style="text-align: justify;">
<h1 style="background-color: #f0f0f0;">Method</h1>
<p style="text-align: justify; background-color: #f0f0f0">
We present a novel unified model, i.e., <a style="color: #70a8dc;">S</a><a style="color: #6fb051;">h</a><a style="color: #e06766;">o</a><a style="color: #f7b26b;">w</a>-o,
capable of addressing both multimodal understanding and generation tasks simultaneously with mixed auto-regressive and diffusion modeling.
</p>
Expand All @@ -230,27 +226,27 @@ <h1>Method</h1>

<div class="album py-5 bg-body-tertiary text-dark ">

<div class="container col-lg-6 mx-auto">
<div class="container col-lg-6 mx-auto" style="background-color: #f0f0f0;">

<h1>Text-to-Image Results</h1>
<img src="./assets/images/github_t2i.png" style="width: 100%;">
<h1 style="background-color: #f0f0f0;">Text-to-Image Results</h1>
<img src="./assets/images/github_t2i.png" style="width: 100%;background-color: #f0f0f0;">

<h1>Multimodal Understanding Results</h1>
<img src="./assets/images/github_mmu.png" style="width: 100%;">
<h1 style="background-color: #f0f0f0;">Multimodal Understanding Results</h1>
<img src="./assets/images/github_mmu.png" style="width: 100%;background-color: #f0f0f0;">

<h1>Inpainting Results</h1>
<img src="./assets/images/github_inpainting.png" style="width: 100%;">
<h1>Extrapolation Results</h1>
<h1 style="background-color: #f0f0f0;">Inpainting Results</h1>
<img src="./assets/images/github_inpainting.png" style="width: 100%;background-color: #f0f0f0;">
<h1 style="background-color: #f0f0f0;">Extrapolation Results</h1>

<img src="./assets/images/github_extrapolation.png" style="width: 100%;">
<img src="./assets/images/github_extrapolation.png" style="width: 100%; background-color: #f0f0f0;">
<!-- <img src="./assets/images/understanding_result.png" style="width: 100%;"> -->
</div>
</div>

<div class="album py-5 bg-body-tertiary text-dark">
<div class="container col-lg-6 mx-auto">
<div class="container col-lg-6 mx-auto" style="background-color: #f0f0f0;">

<h1>Experiments</h1>
<h1 style="background-color: #f0f0f0;">Experiments</h1>
<img src="./assets/images/understanding.png" style="width: 100%;">
<img src="./assets/images/geneval_result.png" style="width: 100%;">

Expand All @@ -260,9 +256,9 @@ <h1>Experiments</h1>

<div class="album py-5 bg-body-tertiary text-dark ">

<div class="container col-lg-6 mx-auto">
<div class="container col-lg-6 mx-auto" style="background-color: #f0f0f0;">

<h1>Comparison</h1>
<h1 style="background-color: #f0f0f0;">Comparison</h1>

<img src="./assets/images/generation_result.png" style="width: 100%;">
<img src="./assets/images/understanding_result.png" style="width: 100%;">
Expand Down

0 comments on commit 783bea7

Please sign in to comment.