fix(index): update figs

Author: ximingxing

assets/img/abl_acc-1.png

@@ -120,9 +120,9 @@ <h3>Technical Report</h3>
                 <div class="col col-lg-2">
                     <a id="paper" href="#">Paper</a>
                 </div>
-                <div class="col col-lg-2">
-                    <a id="arxiv" href="#">Arxiv</a>
-                </div>
+                <!-- <div class="col col-lg-2"> -->
+                    <!-- <a id="arxiv" href="#">Arxiv</a> -->
+                <!-- </div> -->
                 <div class="col col-lg-2">
                     <a id="code" href="#">Code</a>
                 </div>
@@ -138,7 +138,11 @@ <h3>Technical Report</h3>
         <figure class="figure">
             <img src="assets/img/teaser-1.png" class="img-thumbnail figure-img img-fluid rounded"
                  alt="beautiful teaser">
-            <figcaption class="figure-caption">High quality images generated by HunyuanImage 2.1 with <b>PromptEnhancer</b></figcaption>
+            <figcaption class="figure-caption">
+                <b>PromptEnhancer enables high-fidelity and stylistically diverse image generation from user prompts.</b> 
+                Using HunyuanImage 2.1 as the base T2I model, our method demonstrates its versatility across various domains, including photorealism, digital art, abstract geometry, and multilingual text-in-image generation. 
+                The examples showcase how minimal user inputs are transformed into rich, detailed prompts that yield high-quality visual outputs, bridging the gap between user intent and model execution.
+            </figcaption>
         </figure>
     </div>
 
@@ -265,13 +269,28 @@ <h2 class="card-title card-title-margin">Data Pipeline</h2>
                 <figure class="figure">
                     <img src="assets/img/pipe_data-1.png" class="figure-img img-fluid rounded" alt="method">
                     <figcaption class="figure-caption">
-                        Overview of the construction and filtering pipeline for the Rewriter training data. 
+                        <b>Overview of the construction and filtering pipeline for the Rewriter training data.</b> 
+                        The process involves user prompt simulation, Gemini-based generation, human-in-the-loop selection, and automated filtering to ensure high quality.
+                    </figcaption>
+                </figure>
+            </div>
+        </div>
+
+        <div class="card-body">
+            <h2 class="card-title card-title-margin">Data Analysis</h2>
+            <hr class="body-hr"/>
+            <div class="text-center">
+                <figure class="figure">
+                    <img src="assets/img/dataset_analyst-1.png" class="figure-img img-fluid rounded" alt="method">
+                    <figcaption class="figure-caption">
+                        <b>Distribution of evaluation dimensions in our dataset.</b>
+                        (a) The detailed percentage of each of the 24 fine-grained KeyPoints, sorted in descending order. 
+                        (b) The aggregated percentage for each of the six main Super-Categories, calculated by summing the percentages of their constituent KeyPoints. In both charts, colors represent the Super-Category, visually linking the detailed points to their broader classification.
                     </figcaption>
                 </figure>
             </div>
         </div>
     </div>
-    <br>
 
     <div class="card">
         <div class="card-body">
@@ -357,11 +376,11 @@ <h2 class="card-title card-title-margin">Supplementary Material</h2>
         <div class="card-body">
             <h2 class="card-title card-title-margin">Citation</h2>
             <p id="bibtex" class="card-text" style="border: #1a1c20 1px">
-                <code> @article{author20XXtitle,<br>
-                    &nbsp;&nbsp;title={Please cite me},<br>
-                    &nbsp;&nbsp;author={Author, First and Author, Second and Author, Third},<br>
+                <code> @article{promptenhancer2025,<br>
+                    &nbsp;&nbsp;title={PromptEnhancer: A Simple Approach to Enhance Text-to-Image Models via Chain-of-Thought Prompt Rewriting},<br>
+                    &nbsp;&nbsp;author={Linqing Wang and Ximing Xing and Yiji Cheng and Zhiyuan Zhao and Jiale Tao and QiXun Wang and Ruihuang Li and Xin Li and Mingrui Wu and Xinchi Deng and Chunyu Wang and Qinglin Lu},<br>
                     &nbsp;&nbsp;booktitle={arXiv preprint},<br>
-                    &nbsp;&nbsp;year={20XX}<br>
+                    &nbsp;&nbsp;year={2025}<br>
                     }
                 </code>
             </p>