To access this work you must either be on the Smith College campus OR have valid Smith login credentials.
On Campus users: To access this work if you are on campus please Select the Download button.
Off Campus users: To access this work from off campus, please select the Off-Campus button and enter your Smith username and password when prompted.
Non-Smith users: You may request this item through Interlibrary Loan at your own library.
Publication Date
2023-5
First Advisor
Nicholas R. Howe
Document Type
Honors Project
Degree Name
Bachelor of Arts
Department
Computer Science
Keywords
handwriting identification, self-supervised learning, handwriting style analysis, ancient Syriac document identification, contrastive representation learning
Abstract
This thesis presents a series of methods of representation learning for ancient Syriac manuscripts style identification. Our research started with grapheme-based method that reduces the dimension of the images with a codebook trained by k-means clustering and then moved to traditional supervised learning with fine-tuning the pre-trained VGG model. We then explored with contrastive representation learning framework with our dataset and found its applicability in handwriting-based images. Furthermore, we proposed a new self-supervised style content separation model that extracts the style information and uses them to classify the manuscripts. Our model performs well in the document identification task and shows strong potential in manuscript dating. Our work provides insight into utilizing self-supervised learning to find a representation space where manuscripts sharing similar styles would cluster while those in different styles would be away from each other.
Rights
©2023 Yanning Tan. Access limited to the Smith College community and other researchers while on campus. Smith College community members also may access from off-campus using a Smith College log-in. Other off-campus researchers may request a copy through Interlibrary Loan for personal use.
Language
English
Recommended Citation
Tan, Yanning, "Self-supervised Learning for Ancient Syriac Handwriting Style Identification" (2023). Honors Project, Smith College, Northampton, MA.
https://scholarworks.smith.edu/theses/2548
Smith Only:
Off Campus Download