Update README.md

README.md

@@ -45,7 +45,7 @@ As generative AI (like SDXL, DALL·E, and Midjourney) becomes more accessible, t
 
 ## Model Overview  
 
-**SDXL-Deepfake-Detector** is a fine-tuned vision transformer that classifies human faces as **artificial (0)** or **human (1)**, achieving an accuracy of **86%**.
+**SDXL-Deepfake-Detector** is a fine-tuned vision transformer that classifies human faces as **artificial (0)** or **human (1)**, achieving an accuracy of **97%**.
 
 ## Training Approach