Handwriting recognition refers to interpreting and analyzing handwritten text. In rece-nt years, there have- been notable advance-ments in this field, espe-cially in the context of computerize-d assessments. As online e-xams and digital education platforms continue to gain popularity, handwriting recognition plays a crucial role- in evaluating students' written answers. Our proposed system automatically recognizes and scores handwritten responses on answer sheets by comparing them to the correct answers provided by a moderator. To achieve this, the system utilizes Optical Character Recognition (OCR) to convert the handwritten text images into computer-readable text. Additionally, BERT is employed to convert the text into embeddings, and cosine similarity is utilized to take these embeddings as input and provide a final matching confidence score.

Keywords : OCR, Google Vision OCR, BERT, Cosine similarity.