Merge pull request #1 from SALT-NLP/XenonMolecule-patch-1

Update Speech Arena to Talk Arena
SALT-NLP · Dec 2, 2024 · 1ee4c63 · 1ee4c63
2 parents b64d3af + b4b7f58
commit 1ee4c63
Showing 1 changed file with 3 additions and 3 deletions.
diff --git a/index.html b/index.html
@@ -3,7 +3,7 @@
   <head>
     <meta charset="utf-8">
     <title>Talk Arena</title>
-    <meta name="description" content="Distilling an End-to-End Voice Assistant from Speech Recognition Data Using Pretrained Models.">
+    <meta name="description" content="Talk Arena: Interactive Evaluation of Large Audio Models">
 
     <meta charset="UTF-8">
     <meta name="viewport" content="width=device-width, initial-scale=1.0">
@@ -124,7 +124,7 @@ <h2>Interactive Evaluation of Large Audio Models</h2>
 		<span class="author-block">
       <a href="https://yocodeyo.github.io/">Ella Minzhi Li</a><sup><span title="Project Lead, Stanford University & National University of Singapore" alt="Project Lead, Stanford University & National University of Singapore">* †</span></sup>,</span>
 		<span class="author-block"><a href="https://williamheld.com/">Will Held</a><sup><span title="Evaluation Co-Author, Georgia Institute of Technology & Stanford University" alt="Evaluation Coauthor, Georgia Institute of Technology & Stanford University">* †<span></sup>,</span>
-		  <span class="author-block"><a href="https://michryan.com/">Michael Ryan</a><sup><span title="Evaluation Co-Author, Stanford University" alt="Evaluation Coauthor, Stanford University">†<span></sup>,</span>
+		  <span class="author-block"><a href="https://michryan.com/">Michael J. Ryan</a><sup><span title="Evaluation Co-Author, Stanford University" alt="Evaluation Coauthor, Stanford University">†<span></sup>,</span>
 		    <span class="author-block"><a href="https://www.zhuhao.me/">Hao Zhu</a><sup><span title="Evaluation Co-Author, Stanford University" alt="Evaluation Coauthor, Stanford University">†<span></sup>,</span>
 		      <span class="author-block">
 			  <a href="https://cs.stanford.edu/~diyiy/">Diyi Yang</a><sup><span alt="Project Advisor, Stanford University" title="Project Advisor, Stanford University" style="z-index: 1000;">** </span></sup>
@@ -226,7 +226,7 @@ <h2 class="title is-3">Overview</h2>
 
 		<br/><br/>
 
-		Recent efforts towards creating multimodal models have resulted in LLMs capable of processing audio inputs such as speech.  Speech is a low-friction interface which expands social and phonetic interaction opportunities with end users.  Prior work has benchmarked audio models on a set of disjoint static audio tests such as sarcasm or humor detection.  However such static benchmarks lack the complex dynamics of real user interactions and preferences.  Inspired by arena-style evaluations for text LLMs we introduce Speech Arena, an open platform for evaluating Large Audio Models with pairwise human preferences.  Speech Arena helps to reveal insights on:
+		Recent efforts towards creating multimodal models have resulted in LLMs capable of processing audio inputs such as speech.  Speech is a low-friction interface which expands social and phonetic interaction opportunities with end users.  Prior work has benchmarked audio models on a set of disjoint static audio tests such as sarcasm or humor detection.  However such static benchmarks lack the complex dynamics of real user interactions and preferences.  Inspired by arena-style evaluations for text LLMs we introduce Talk Arena, an open platform for evaluating Large Audio Models with pairwise human preferences.  Talk Arena helps to reveal insights on:
                     <br><br>
                     <b>What use cases users are exploring with large audio models?</b> We can analyze user queries from the wild and compare the use case difference with traditional use cases of text LLMs.
                     <br><br>