Skip to content

Commit

Permalink
change page format
Browse files Browse the repository at this point in the history
  • Loading branch information
@francescotaioli
francescotaioli committed Dec 2, 2024
1 parent 4f16bfc commit 267cb67
Showing 1 changed file with 80 additions and 13 deletions.
93 changes: 80 additions & 13 deletions index.html
View file
Original file line number Diff line number Diff line change
Expand Up @@ -98,37 +98,50 @@ <h1 class="title is-1 publication-title">
</h1>
<div class="is-size-5 publication-authors">
<span class="author-block">
<a href="https://francescotaioli.github.io/"
<a href="https://francescotaioli.github.io/" target="_blank"
>Francesco Taioli</a
><sup>1,2</sup>,</span
>
<span class="author-block">
<a
href="https://scholar.google.com/citations?hl=it&user=fqdv3d4AAAAJ&view_op=list_works&sortby=pubdate"
target="_blank"
>Edoardo Zorzi</a
><sup>2</sup>,
</span>
<span class="author-block"
><a href="https://giannifranchi.github.io/">Gianni Franchi</a
><a
href="https://scholar.google.it/citations?hl=it&newwindow=1&user=ZCW6-psAAAAJ&view_op=list_works&sortby=pubdate"
target="_blank"
>Gianni Franchi</a
><sup>3</sup>,
</span>
<span class="author-block">
<a href="https://profs.scienze.univr.it/~castellini/"
<a
href="https://scholar.google.it/citations?hl=it&user=AploBScAAAAJ&view_op=list_works&sortby=pubdate"
target="_blank"
>Alberto Castellini</a
><sup>2</sup>,
</span>
<span class="author-block">
<a href="http://profs.sci.univr.it/~farinelli/"
<a
href="https://scholar.google.co.uk/citations?hl=en&user=KHAIAA8AAAAJ&view_op=list_works&sortby=pubdate"
target="_blank"
>Alessandro Farinelli</a
><sup>2</sup>,
</span>
<span class="author-block"
><a href="https://www.dimi.univr.it/?ent=persona&id=218"
><a
href="https://scholar.google.com/citations?hl=en&user=LbgTPRwAAAAJ&view_op=list_works&sortby=pubdate"
target="_blank"
>Marco Cristani</a
><sup>2</sup>,
</span>
<span class="author-block"
><a href="https://www.yimingwang.it/">Yiming Wang</a
><a
href="https://scholar.google.co.uk/citations?hl=en&user=KBZ3zrEAAAAJ&view_op=list_works&sortby=pubdate"
target="_blank"
>Yiming Wang</a
><sup>4</sup>
</span>
</div>
Expand Down Expand Up @@ -205,7 +218,7 @@ <h1 class="title is-1 publication-title">
alt="Teaser"
/>
<div class="content has-text-justified">
<p style="padding: 0px 2em 0 2em">
<p style="padding: 1em 2em 0em 2em">
Sketched episode of the proposed
<b><i>Collaborative Instance Navigation (CoIN)</i></b>
task. The human user (bottom left) provides a request (<i
Expand Down Expand Up @@ -233,7 +246,7 @@ <h1 class="title is-1 publication-title">
></b
>, producing a refined
<b><font color="green">detection description</font></b
>. The<b>Interaction Trigger</b> uses this refined
>. The <b>Interaction Trigger</b> uses this refined
description to decide whether to pose a question to the
user (①,③,④), continue the navigation (②) or halt the
exploration (⑤).
Expand All @@ -247,11 +260,65 @@ <h1 class="title is-1 publication-title">
</div>
</section>

<section class="hero teaser">
<div class="container">
<div class="columns is-centered">
<div class="column is-max-desktop">
<div class="content has-text-centered">
<div class="columns is-centered">
<div class="column is-10">
<br />
<h2 class="title is-3">Abstract</h2>
<div class="content has-text-justified">
<p style="padding: 1em 2em 0em 2em">
Existing embodied instance goal navigation tasks, driven
by natural language, assume human users to provide
complete and nuanced instance descriptions prior to the
navigation, which can be impractical in the real world as
human instructions might be brief and ambiguous.
</p>
<p style="padding: 0em 2em 0em 2em">
&nbsp;&nbsp;&nbsp;To bridge this gap, we propose a new
task, Collaborative Instance Navigation (CoIN), with
dynamic agent-human interaction during navigation to
actively resolve uncertainties about the target instance
in natural, template-free, open-ended dialogues.
</p>
<p style="padding: 0em 2em 0em 2em">
&nbsp;&nbsp;&nbsp;To address CoIN, we propose a novel
method, Agent-user Interaction with UncerTainty Awareness
(AIUTA), leveraging the perception capability of Vision
Language Models (VLMs) and the capability of Large
Language Models (LLMs). First, upon object detection, a
Self-Questioner model initiates a self-dialogue to obtain
a complete and accurate observation description, while a
novel uncertainty estimation technique mitigates
inaccurate VLM perception. Then, an Interaction Trigger
module determines whether to ask a question to the user,
continue or halt navigation, minimizing user input.
</p>

<p style="padding: 0em 2em 0em 2em">
&nbsp;&nbsp;&nbsp;For evaluation, we introduce CoIN-Bench,
a benchmark supporting both real and simulated humans.
AIUTA achieves competitive performance in instance
navigation against state-of-the-art methods, demonstrating
great flexibility in handling user inputs.
</p>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</section>

<section class="section">
<div class="container is-max-desktop">
<!-- Abstract. -->
<div class="container is-max-desktop is-centered">
<!-- Abstract.
<div class="columns is-centered has-text-centered">
<div class="column is-full">
<div class="column is-10">
<h2 class="title is-3">Abstract</h2>
<div class="content has-text-justified">
<p>
Expand Down Expand Up @@ -290,13 +357,13 @@ <h2 class="title is-3">Abstract</h2>
</p>
</div>
</div>
</div>
</div> -->
<!--/ Abstract. -->

<!-- Paper video. -->
<div class="columns is-centered has-text-centered">
<div class="column is-is-full">
<h2 class="title is-3">Video</h2>
<h2 class="title is-3">Method</h2>
<div class="publication-video">
<iframe
style="display: block; background-color: white"
Expand Down

0 comments on commit 267cb67

Please sign in to comment.