Update README.md

README.md

@@ -16,7 +16,7 @@ Official implementation of the paper: **"Vision-Language Subspace Prompting"**.
 
 ## ✨ Introduction
 <div align="left">
-  <img src="docs/insight.jpg" width="50%">
+  <img src="docs/insight.jpg" width="80%">
 </div>
 <!-- ![Main Insight](docs/insight.jpg) -->
 
@@ -68,7 +68,7 @@ SuPr's subspace modeling captures diverse intra-class variations, including fine
 ![Subspace Visualization](docs/vis.jpg)
 
 <div align="left">
-  <img src="docs/walking.jpg" width="80%">
+  <img src="docs/walking.jpg" width="50%">
 </div>
 
 ---