zodumair/document-forgery-detector

Document Forgery Detector

A Vision Transformer (ViT) based document forgery detection system using Error Level Analysis (ELA).

Final Year Project — Sir Syed University of Engineering & Technology, Karachi

• —91.3% accuracy on test set
• —Detects: copy-move, text replacement, stamp overlay, compression artifacts, splicing
• —Base model: google/vit-base-patch16-224