Landmark/key point image annotation for document authentication (AI Engineer Intern)
I developed a YOLOv11-based landmark detection model to localize security features and critical fields on document images. The work involved custom key point annotation of eKYC datasets followed by iterative model evaluation. This enabled automated region extraction for subsequent document verification workflows. • Annotated key landmarks and reference points on identity document images • Labeled regions of interest for automated alignment and extraction • Benchmarked key point detection accuracy using labeled datasets • Created annotation documentation for reproducibility and quality control