add gif

index.html

@@ -30,10 +30,14 @@ <h5>ICML 2024</h5>
         </p>
   </font>
   <br>
-  <img src="./docs/teaser.png" class="teaser-gif" style="width:100%;"><br>
+  <img src="./docs/teaser.jpg" class="teaser-gif" style="width:100%;"><br>
   <h4 style="text-align:center"><em>Multi-Concept Prompt Learning (MCPL) pioneers the novel task of mask-free text-guided learning for multiple prompts from one scene. Our approach not only enhances current methodologies but also paves the way for novel applications, such as facilitating knowledge discovery through natural language-driven interactions between humans and machines. </em></h4>
 </div>
 
+<div class="content">
+  <img style="width: 100%;" src="./docs/teaser.gif" alt="teaser.">
+</div>
+
 <div class="content">
   <h2 style="text-align:center;">Abstract</h2>
   <p>Textural Inversion, a prompt learning method, learns a singular embedding for a new "word" to represent image style and appearance, allowing it to be integrated into natural language sentences to generate novel synthesised images. However, identifying and integrating multiple object-level concepts within one scene poses significant challenges even when embeddings for individual concepts are attainable. This is further confirmed by our empirical tests. To address this challenge, we introduce a framework for <i>Multi-Concept Prompt Learning (MCPL)</i>, where multiple new "words" are simultaneously learned from a single sentence-image pair. To enhance the accuracy of word-concept correlation, we propose three regularization techniques: <i>Attention Masking (AttnMask)</i> to concentrate learning on relevant areas; <i>Prompts Contrastive Loss (PromptCL)</i> to separate the embeddings of different concepts; and <i>Bind adjective (Bind adj.)</i> to associate new "words" with known words. We evaluate via image generation, editing, and attention visualization with diverse images. Extensive quantitative comparisons demonstrate that our method can learn more semantically disentangled concepts with enhanced word-concept correlation. Additionally, we introduce a novel dataset and evaluation protocol tailored for this new task of learning object-level concepts.</p>