Update README.md

README.md

@@ -1,3 +1,13 @@
+---
+license: mit
+title: VisionSort
+sdk: streamlit
+emoji: 💻
+colorFrom: gray
+colorTo: indigo
+pinned: true
+short_description: 'AI visual search tool for images using natural language '
+---
 # VisionSort  
 *AI-powered visual search tool for finding key moments in large batches of images or video frames using natural prompts.*
 
@@ -91,7 +101,4 @@ The following features were previously implemented but later removed  (archived
   - Managing multiple file uploads and temp files  
   - Keeping UI responsive with real-time feedback (match scores, timestamps, toggles)
 - Key takeaway: **Build a stable core first**, then layer in advanced features.
-- Although GPT-4 and BLIP weren’t fully integrated in the final version, I preserved and documented their experiments for future improvements.
-
-
-
+- Although GPT-4 and BLIP weren’t fully integrated in the final version, I preserved and documented their experiments for future improvements.