Edvanced, exams management tool

Developed an end‑to‑end system to extract and analyze answers from scanned bubble sheets using OpenCV, OCR,

and ArUco markers, enabling accurate identification of student responses and IDs

• Implemented a transparent and explainable cheating detection pipeline by combining statistical analysis with clus‑

tering algorithms (DBSCAN + PCA), flagging suspicious patterns using p‑value assessments and visual cluster insights

• Built a Streamlit app allowing users to upload exam sheets, visualize scores, review results and suspected cheaters