clean text

app.py

@@ -14,14 +14,20 @@ learned_emojis_base64 = ''
 with open(f'{base_path}/data/learned_emojis.png', 'rb') as f:
     learned_emojis_base64 = base64.b64encode(f.read()).decode()
 example_images = [f'{base_path}/data/test_images/smile.png']
-description = '''
-This project tries to classify emojis.  
-The model was only trained on outline images, while the input is now hand drawn.  
-See how it was created: https://www.kaggle.com/code/krayc81/emoji-doodle
-'''
+description = ''
 article = f'''
+<h3>Description</h3>
+This project focuses on classifying hand-drawn emojis.  
+Interestingly, the model was trained exclusively on outline images, yet it's capable of interpreting hand-drawn input.  
+This approach demonstrates the model's ability to generalize from simplified representations to more varied, real-world examples.  
+Curious about the development process? You can explore the project's creation in detail here:  
+<a href="https://www.kaggle.com/code/krayc81/emoji-doodle">Click here</a>
 <h3>How to Use:</h3>
-Try to draw one of the following emojis and click submit when you are ready
+1. Choose one of the emojis in the picture underneath<br>
+2. Using the drawing interface on the left, sketch your interpretation of the chosen emoji.<br>
+3. When you are satisfied with your drawing, click the "Submit" button.<br><br>
+Your hand-drawn emoji will then be analyzed by our classification model.<br>
+This interactive process allows you to test the model's ability to recognize various emoji styles and interpretations.<br>
 <img src="data:image/png;base64,{learned_emojis_base64}">
 '''