top of page

SEM-LEX

SEM-LEX is a large collection of videos of individual signs in ASL, with over 84,000 videos of more than 3,100 signs produced by deaf signers. The videos come from a word-association task: signers saw a sign and responded with the first signs related in meaning that came to mind. That makes SEM-LEX useful in two ways. It shows how signs are related to one another in terms of meaning (i.e., cat is semantically related to dog, butter is related to toast etc.). The dataset is also useful for building technology that can recognize individual signs.

The Team

SEM-LEX is a collaboration with the Cognition, Language and Plasticity Lab at Chapman University, and the Laboratory for Language and Cognitive Neuroscience at San Diego State University.

White woman with blonde hair

Zed Sevcikova Sehyr

Chapman University

White woman with blonde hair

Karen Emmorey

San Diego State University

White man with brown hair

Lee Kezar

University of Southern California

White man with blonde hair

Jesse Thomason

University of Southern California

AP1GczMusKKyVahWw0pOZngz-7bEsJVcEQ1V3atc2Cjqa2g9aFiuWqe4O8O4nlAvFBrOUDXgMbImtN-cwdBjHLCxBB

Elana Pontecorvo

Boston University

White woman with curly brown hair

Ruth Ferster

Boston University

White man with brown hair

Connor Baer

Boston University

White woman with curly brown hair

Adele Daniels

Boston University

White woman with blonde hair

Lauren Berger

Boston University

White woman with brown hair

Naomi Caselli

Boston University

Want to replicate SEM-LEX for your sign language?

We are always eager to collaborate.

Thanks for submitting!

bottom of page