Skip to content

Commit

Permalink
modify:index
Browse files Browse the repository at this point in the history
  • Loading branch information
JinSY515 committed Mar 27, 2024
1 parent 238e5c6 commit d2f8c92
Showing 1 changed file with 5 additions and 5 deletions.
10 changes: 5 additions & 5 deletions index.html
Original file line number Diff line number Diff line change
Expand Up @@ -51,7 +51,7 @@ <h1 class="title is-3 publication-title">MoDiTalker: Motion-Disentangled Diffusi
Jiyoung Kim<sup>1</sup></a>,
</span>
<span class="author-block">
<a href=""></a>Jisu Nam<sup>1</sup></a>,
<a href="">Jisu Nam<sup>1</sup></a>,
</span>
<span class="author-block">
<a href="https://cvlab.korea.ac.kr/members/faculty">Seungryong Kim<sup>&dagger;1</sup></a>,
Expand Down Expand Up @@ -195,7 +195,7 @@ <h2 class="title is-3">Talking Head Generation Results</h2>
<div class="content has-text-justified">
<p style="font-size: 15px;">
<strong>Self-reconstruction results of MoDiTalker</strong>
Results of MoDiTalker in self-reconstruction setting, in which there exists the reference video.
Results in self-reconstruction setting, in which exists the reference video.
<div>
<div style="float:left; width: 33%;">
<video controls width="200">
Expand All @@ -221,7 +221,7 @@ <h2 class="title is-3">Talking Head Generation Results</h2>
<div class="content has-text-justified">
<p style="font-size: 15px;">
<strong>One-shot setting results of MoDiTalker</strong>
Results of the MoDiTalker in one-shot setting, in which the reference image is only the first frame of the video.
Results of in one-shot setting, in which the reference image is the single frame of the video.
<div>
<div style="float:left; width: 33%;">
<video controls width="200">
Expand All @@ -247,7 +247,7 @@ <h2 class="title is-3">Talking Head Generation Results</h2>
<div class="content has-text-justified">
<p style="font-size: 15px;">
<strong>Cross-id results of MoDiTalker</strong>
Results of MoDiTalker in cross-id setting, in which the identity of audio and reference frames are different. Those results are generated for the same audio, with different identity.
Results in cross-id setting, in which the identity of audio and reference frames are different. Those results are generated for the same audio, with different identity.
<div>
<div style="float:left; width: 33%;">
<video controls width="200">
Expand All @@ -273,7 +273,7 @@ <h2 class="title is-3">Talking Head Generation Results</h2>
<div class="content has-text-justified">
<p style="font-size: 15px;">
<strong>Out-of-Distribution results of MoDiTalker</strong>
Results of MoDiTalker in cross-id setting, in which the audio is actually a song. The left video is the result of the first stage, AToM, and the right video is the final result of MoDiTalker.
Results in out-of-distribution cross-id setting. The left video is the generated result of the first stage, AToM, and the right video is the final result of MoDiTalker.
<div>
<div style="float:left; width: 33%;">
<video controls width="200">
Expand Down

0 comments on commit d2f8c92

Please sign in to comment.