This thesis focuses on developing automated deep learning methods for writer identification and writer attribute prediction from handwriting. It introduces two novel architectures: Convolutional Transformer Encoder (CTE) and Convolutional Swin Encoder (CSE). CTE is designed for determining authorship from handwritten text, be it modern or historical handwriting. CTE tracks subtle features and cues from handwriting strokes to distinguish authorship. It is the first of its kind capable to operate on historical handwritten fragments, setting it apart from existing methods that rely on entire document pages. CSE is designed to determine multiple attributes of an author such as authorship, gender, age and handedness. It achieves competitive performance compared to traditional page-level methods that typically rely on separate classifiers. While CTE is an exclusive classifier using a Transformer block, CSE is a multi-label classifier leveraging Swin Transformer blocks.
","abstract_html":"<p>This thesis focuses on developing automated deep learning methods for writer identification and writer attribute prediction from handwriting. It introduces two novel architectures: Convolutional Transformer Encoder (CTE) and Convolutional Swin Encoder (CSE). CTE is designed for determining authorship from handwritten text, be it modern or historical handwriting. CTE tracks subtle features and cues from handwriting strokes to distinguish authorship. It is the first of its kind capable to operate on historical handwritten fragments, setting it apart from existing methods that rely on entire document pages. CSE is designed to determine multiple attributes of an author such as authorship, gender, age and handedness. It achieves competitive performance compared to traditional page-level methods that typically rely on separate classifiers. While CTE is an exclusive classifier using a Transformer block, CSE is a multi-label classifier leveraging Swin Transformer blocks.</p>","abstract_has_math":false,"creators":["Majithia, Aditya"],"institution":null,"degree_name":"Master of Science (M.S.)","degree_level":"Thesis","degree_discipline":"Computer Science","degree_department":null,"school":null,"contributors":["Michael Grossberg"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2025,"date_issued":"2025-01-01T08:00:00Z","date_published":"2025-01-01T08:00:00Z","updated_at":"2026-07-24T01:58:13Z","subjects":["Deep Learning","Writer identification","Writer attributes","Individual word images","Historical fragments analysis","Multi-label classification.","Data Science"],"languages":[],"rights":[],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://academicworks.cuny.edu/cc_etds_theses/1270","outbound_label":"Repository record","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Michael Grossberg"]},{"key":"dc:creator","label":"Author","values":["Majithia, Aditya"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.available","label":"Dc Date Available","values":["2025-05-23T07:00:00Z"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Computer Science"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Thesis"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Master of Science (M.S.)"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Deep Learning","Writer identification","Writer attributes","Individual word images","Historical fragments analysis","Multi-label classification.","Data Science"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://academicworks.cuny.edu/cc_etds_theses/1270"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["This thesis focuses on developing automated deep learning methods for writer identification and writer attribute prediction from handwriting. It introduces two novel architectures: Convolutional Transformer Encoder (CTE) and Convolutional Swin Encoder (CSE). CTE is designed for determining authorship from handwritten text, be it modern or historical handwriting. CTE tracks subtle features and cues from handwriting strokes to distinguish authorship. It is the first of its kind capable to operate on historical handwritten fragments, setting it apart from existing methods that rely on entire document pages. CSE is designed to determine multiple attributes of an author such as authorship, gender, age and handedness. It achieves competitive performance compared to traditional page-level methods that typically rely on separate classifiers. While CTE is an exclusive classifier using a Transformer block, CSE is a multi-label classifier leveraging Swin Transformer blocks.
"]},{"key":"dc:title","label":"Title","values":["Using Deep Neural Nets in Writer Identification & Analysis"]}]}],"canonical_facts":{"dc:contributor":["Michael Grossberg"],"dc:creator":["Majithia, Aditya"],"dc:date.available":["2025-05-23T07:00:00Z"],"dc:description.abstract":["This thesis focuses on developing automated deep learning methods for writer identification and writer attribute prediction from handwriting. It introduces two novel architectures: Convolutional Transformer Encoder (CTE) and Convolutional Swin Encoder (CSE). CTE is designed for determining authorship from handwritten text, be it modern or historical handwriting. CTE tracks subtle features and cues from handwriting strokes to distinguish authorship. It is the first of its kind capable to operate on historical handwritten fragments, setting it apart from existing methods that rely on entire document pages. CSE is designed to determine multiple attributes of an author such as authorship, gender, age and handedness. It achieves competitive performance compared to traditional page-level methods that typically rely on separate classifiers. While CTE is an exclusive classifier using a Transformer block, CSE is a multi-label classifier leveraging Swin Transformer blocks.
"],"dc:identifier":["https://academicworks.cuny.edu/cc_etds_theses/1270"],"dc:subject":["Deep Learning","Writer identification","Writer attributes","Individual word images","Historical fragments analysis","Multi-label classification.","Data Science"],"dc:title":["Using Deep Neural Nets in Writer Identification & Analysis"],"thesis:degree_discipline":["Computer Science"],"thesis:degree_level":["Thesis"],"thesis:degree_name":["Master of Science (M.S.)"]},"updated_at":"2026-07-24T01:58:13Z"}