update matchanything

zju3dv · Jan 9, 2025 · 2a3b72d · 2a3b72d
1 parent 6c40804
commit 2a3b72d
Show file tree

Hide file tree

Showing 15 changed files with 39 additions and 9 deletions.
diff --git a/matchanything/css/bootstrap-4.4.1.css → MatchAnything/css/bootstrap-4.4.1.css b/matchanything/css/bootstrap-4.4.1.css → MatchAnything/css/bootstrap-4.4.1.css
diff --git a/matchanything/img/proj_page_a.png → MatchAnything/img/proj_page_a.png b/matchanything/img/proj_page_a.png → MatchAnything/img/proj_page_a.png
diff --git a/matchanything/img/proj_page_b.png → MatchAnything/img/proj_page_b.png b/matchanything/img/proj_page_b.png → MatchAnything/img/proj_page_b.png
diff --git a/matchanything/img/results_a.png → MatchAnything/img/results_a.png b/matchanything/img/results_a.png → MatchAnything/img/results_a.png
diff --git a/matchanything/img/results_b.png → MatchAnything/img/results_b.png b/matchanything/img/results_b.png → MatchAnything/img/results_b.png
diff --git a/matchanything/img/results_c.png → MatchAnything/img/results_c.png b/matchanything/img/results_c.png → MatchAnything/img/results_c.png
diff --git a/matchanything/img/spinner.svg → MatchAnything/img/spinner.svg b/matchanything/img/spinner.svg → MatchAnything/img/spinner.svg
diff --git a/matchanything/img/statistics.png → MatchAnything/img/statistics.png b/matchanything/img/statistics.png → MatchAnything/img/statistics.png
diff --git a/matchanything/img/teaser_v0.mov → MatchAnything/img/teaser_v0.mov b/matchanything/img/teaser_v0.mov → MatchAnything/img/teaser_v0.mov
diff --git a/MatchAnything/img/teaser_video_train_part.png b/MatchAnything/img/teaser_video_train_part.png
diff --git a/matchanything/index.html → MatchAnything/index.html b/matchanything/index.html → MatchAnything/index.html
@@ -262,8 +262,8 @@ <h1 style="font-weight: bold">
             </h1>
             <h4 style="color:#5a6268;">Arxiv 2025</h4>
             <hr>
-            <a href="https://hxy-123.github.io/" target="_blank">Xingyi He<sup>1*</sup> </a> 
-            <a href="https://ritianyu.github.io/" target="_blank">Hao Yu<sup>1,2*</sup> </a> 
+            <a href="https://hxy-123.github.io/" target="_blank">Xingyi He<sup>1</sup> </a> 
+            <a href="https://ritianyu.github.io/" target="_blank">Hao Yu<sup>1,2</sup> </a> 
             <a href="https://pengsida.net" target="_blank">Sida Peng<sup>1</sup> </a> 
             <a href="https://github.com/Cuistiano" target="_blank">Dongli Tan<sup>1</sup> </a> 
             <a href="https://zehongs.github.io" target="_blank">Zehong Shen<sup>1</sup> </a> 
@@ -288,7 +288,7 @@ <h4 style="color:#5a6268;">Arxiv 2025</h4>
                   <a class="btn btn-large btn-light" href="https://github.com/zju3dv/matchanything" role="button"
                     target="_blank">
                     <i class="fa fa-github"></i>
-                    <b>Code</b>
+                    <b>Code (coming soon)</b>
                   </a>
                 </p>
               </div>
@@ -309,8 +309,8 @@ <h3>TL;DR</h3>
             (1) The paper focuses on finding pixel correspondences for image pairs from different imaging principles.<br>
             (2) We propose a large-scale pre-training framework that utilizes <span style="color: #6589bf">synthetic cross-modal</span> training signals, incorporating diverse data from <span
                 style="color: #c68b5b">various sources</span>, to train models to recognize and match fundamental structures across images. This capability is transferable to real-world, unseen cross-modality image matching tasks. <br>
-            (3) The pre-trained models achieve significant boost on various unseen cross-modality image registration tasks, demonstrating strong generalizabilities. <strong>All codes and weights are publicly
-            available.</strong>
+            (3) The pre-trained models achieve significant boost on various unseen cross-modality image registration tasks, demonstrating strong generalizabilities. 
+            <!-- <strong>All codes and weights are publicly available.</strong> -->
           </p>
         </div>
       </div>
@@ -323,11 +323,41 @@ <h3>TL;DR</h3>
         <div class="col-12">
           <h3>Abstract</h3>
           <hr style="margin-top: 0px" />
-          <div class="media-container" style="display: flex; justify-content: center; align-items: center;">
+
+          <head>
+            <meta charset="UTF-8">
+            <meta name="viewport" content="width=device-width, initial-scale=1.0">
+            <title>Image and Video Side by Side</title>
+            <style>
+                .media-section {
+                    display: flex;
+                    align-items: center; /* 保证高度对齐 */
+                    transform: translateX(-6%);
+                }
+                .media-section img,
+                .media-section video {
+                    height: 400px;
+                    width: auto; /* 保持宽高比 */
+                }
+            </style>
+          </head>
+          <body>
+              <div class="media-section">
+                  <!-- 图片 -->
+                  <img src="img/teaser_video_train_part.png" alt="Example Image">
+                  <!-- 视频 -->
+                  <video controls playsinline autoplay loop preload muted>
+                      <source src="img/teaser_v0.mov" type="video/mp4">
+                      Your browser does not support the video tag.
+                  </video>
+              </div>
+          </body>
+
+          <!-- <div class="media-container" style="display: flex; justify-content: center; align-items: center;">
             <video id="demo" width="90%" playsinline autoplay loop preload muted>
               <source src="img/teaser_v0.mov" type="video/mp4">
             </video>
-          </div>
+          </div> -->
 
           <p style="font-size: small;">
             <strong>Capabilities</strong> of the image matching model pre-trained by our framework. Green lines indicate the identified corresponding pixel localizations between images. Using the same network weight, our model exhibits impressive generalization abilities across extensive unseen real-world cross-modality matching tasks, benefiting diverse applications in
@@ -350,7 +380,7 @@ <h3>Abstract</h3>
           <h3>Method</h3>
           <hr style="margin-top: 0px" />
           <div style="display: flex; justify-content: center;">
-            <img src="img/proj_page_a.png" width="90%" />
+            <img src="img/proj_page_a.png" width="95%" />
           </div>
           <p>
             <strong>Preliminary</strong>:
@@ -359,7 +389,7 @@ <h3>Method</h3>
         </div>
         <div class="col-12">
           <div style="display: flex; justify-content: center;">
-            <img src="img/proj_page_b.png" width="90%" />
+            <img src="img/proj_page_b.png" width="95%" />
           </div>
           <p>
             <strong>Pipeline</strong>.

diff --git a/matchanything/js/bootstrap-4.4.1.js → MatchAnything/js/bootstrap-4.4.1.js b/matchanything/js/bootstrap-4.4.1.js → MatchAnything/js/bootstrap-4.4.1.js
diff --git a/matchanything/js/jquery-3.4.1.min.js → MatchAnything/js/jquery-3.4.1.min.js b/matchanything/js/jquery-3.4.1.min.js → MatchAnything/js/jquery-3.4.1.min.js
diff --git a/matchanything/js/popper.min.js → MatchAnything/js/popper.min.js b/matchanything/js/popper.min.js → MatchAnything/js/popper.min.js
diff --git a/matchanything/js/video_comparison.js → MatchAnything/js/video_comparison.js b/matchanything/js/video_comparison.js → MatchAnything/js/video_comparison.js