update homepage

app.py

@@ -56,18 +56,15 @@ div#warning-duplicate .actions a {
 
 with gr.Blocks(css=css) as demo:
     gr.Markdown("# JoyHallo: Digital human model for Mandarin")
-    gr.Markdown("Generate talking head avatars driven with Mandarin speech.")
+    gr.Markdown("Generate talking head avatars driven by Mandarin speech. Data requirements:")
     gr.Markdown("""
-
-Data requirements:
-
 Image:
 1. Cropped to square shape.
-2. Face should be facing forward and occupy 50%-70% of the image area.
+2. Face should be facing forward and occupy 50%-70% of the image.
 
 Audio:
-1. Audio in wav format.
-2. Mandarin or English or mixed, with clear audio and suitable background music.
+1. Use wav format.
+2. Mandarin, English or mixed, with clear audio and suitable background music.
                 
 ! Important: Too long audio will casue a very long processing time, please keep the audio length within 5s.
                 """)