sinhala-print-trocr

A fine-tuned TrOCR model for recognizing printed Sinhala text from images, built on Microsoft's transformer-based OCR architecture. It converts scanned or photographed Sinhala printed documents into machine-readable text strings. Source: supabase.