Update index.html

index.html

@@ -108,7 +108,7 @@
             <div class="columns is-centered">
                <div class="column has-text-centered">
                   <h1 class="title is-1 publication-title" style="margin-bottom:0">
-                     <img src="./static/logo-hao.svg" alt="Talk Arena Logo" style="height: 10vw;"> Talk Arena
+                     <img src="./static/logo-hao.svg" alt="Talk Arena Logo" style="height: 5vw;"> Talk Arena
                   </h1>
                   <h2>Interactive Evaluation of Large Audio Models</h2>
                   <div class="is-size-5 publication-authors">
@@ -157,13 +157,7 @@ <h2>Interactive Evaluation of Large Audio Models</h2>
                   </div>
                   <section class="hero teaser">
                      <div class="container is-max-desktop">
-                        <div class="hero-body">
-                           <div class="columns is-centered has-text-centered">
-                              <div class="container is-max-desktop">
-                                 <gradio-app title="gradio demo" src="https://b06ef30ce0641fa3bd.gradio.live/"></gradio-app>
-                              </div>
-                           </div>
-                           <br/>
+                        <div class="hero-body">	
                            <h4 class="subtitle has-text-centered">
                               <span style="color: #820000">[TL;DR]
                               </span> As AI Assistants' capability expands beyond text into other modalities like audio, they enable new forms of interactions between humans and AI. On the other hand, most benchmarks for audio processing primarily focus on transcribing and analyzing user speech. We introduce <span id="highlight"><b>Talk Arena</b></span>, an interactive open platform to evaluate Large Audio Models through <span id="highlight"><b>interactions with users in real-world settings</b></span>. It helps to assess whether previous static benchmarks are valuable measures of model quality, or whether there is need for a new evaluation paradigm for Large Audio Models. We evaluate <span
@@ -172,6 +166,12 @@ <h4 class="subtitle has-text-centered">
                                  id="highlight"><b>18 speech comprehension datasets</b></span> and <span
                                  id="highlight"><b>Talk Arena</b></span>.</a>
                            </h4>
+                           <br/>
+                           <div class="columns is-centered has-text-centered">
+                              <div class="container is-max-desktop">
+                                 <gradio-app title="gradio demo" src="https://b06ef30ce0641fa3bd.gradio.live/"></gradio-app>
+                              </div>
+                           </div>
                         </div>
                      </div>
                   </section>