![Featured image for “[MS/PhD] Thesis Defense: Phonological inductive biases for computationally modeling American Sign Language video”](https://www.cs.usc.edu/wp-content/uploads/2025/06/USC-Featured-Photo.png)
The following announcement is from [Lee Kezar <lkezar@usc.edu>]. Please contact them directly if you have any questions.
Date: Tuesday, July 15, 9:00 AM
Room: Kaprielian Hall, Room 134
Chair: Jesse Thomason
Title: Phonological inductive biases for computationally modeling American Sign Language video
Abstract:
Computational models for American Sign Language (ASL) are limited by data scarcity and model designs that neglect to consider the language’s internal linguistic structure. This dissertation addresses that gap by systematically integrating this phonological knowledge—the sub-lexical building blocks of signs—as an inductive bias. This defense will review empirical results demonstrating that a neuro-symbolic, knowledge-infused approach improves model accuracy, robustness, and data-efficiency on a wide variety of ASL-input and output tasks.
Published on July 8th, 2025Last updated on July 8th, 2025