Modeling American Sign Language via Linguistic Knowledge Infusion

Explore the challenges and advancements in modeling American Sign Language (ASL) through linguistic knowledge infusion in this informative seminar presented by Lee Kezar from USC. Delve into the innovative approaches addressing data scarcity and low engagement issues in ASL research. Discover how incorporating phonological, morphological, and semantic knowledge can significantly improve sign recognition accuracy, enable few- and zero-shot sign understanding, and reduce demographic biases. Learn about the potential impact of these advancements on increasing access to language technologies for deaf and hard-of-hearing individuals. Gain insights into longstanding research questions in sign language phonology and language acquisition. The presentation includes ASL interpretation, making it accessible to both hearing and deaf audiences.