Date of Award
Summer 8-9-2026
Document Type
Thesis
Degree Name
Master of Science (M.S.)
Department
Computer and Information Sciences
Language
English
First Advisor
Panneer Selvam Santhalingam
Abstract
Video conferencing degrades asymmetrically. When bandwidth falls, a hearing caller loses picture quality and keeps the conversation; a deaf and hard of hearing signer, whose language is carried entirely in the visual modality, loses the conversation. This thesis asks whether signed video reduced to the rates at which commercial platforms fail can be reconstructed at the receiver well enough to keep signing legible. A twostage reduction pipeline crops to the signer and transmits the face and hands at higher fidelity than their surroundings, achieving a reduction of approximately 99%; reconstruction uses a recurrent bottleneck mixer architecture, trained both conventionally and with scheduled sampling adapted from recurrent sequence prediction. Neither model improves on the video it was given. On held-out material both fall several decibels below their own input, which returned to full resolution by interpolation reconstructs the frame better than the enhancement does; scheduled sampling improves on conventional training, but among options none of which clears that floor. The contribution is therefore a characterization rather than a system.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Thantsin, Swann, "Enabling ASL Digital Communication Under Poor Internet Access" (2026). CUNY Academic Works.
https://academicworks.cuny.edu/bc_etds/27
