Immersive VR for Signed Language Learning: A Comparative Evaluation of ASL Vocabulary Production

Limited access to qualified instructors and opportunities for guided practice remains a barrier for many people learning American Sign Language (ASL), including hearing learners and newly signing Deaf people. Interactive technologies that provide real-time, accessible feedback may help lower these barriers and support wider adoption of ASL. We developed ASL Champ!, a standalone virtual reality (VR) application that supports introductory ASL vocabulary learning in immersive environments. ASL Champ! provides automated feedback via a deep learning–based sign recognition algorithm using the VR headset's built-in cameras. We evaluated ASL Champ! in a within-participant experimental study comparing it with a conventional 2D video-based learning platform (Edpuzzle), focusing on user experience, cognitive workload, and behavioral outcomes relevant to accessible learning systems. Adult non-signers ( n = 11) completed matched ASL lessons using ASL Champ! (VR) and Edpuzzle (2D) and completed assessments of receptive ASL comprehension and expressive sign production. Results showed that VR-based learning imposed a higher cognitive workload than the 2D platform, while receptive comprehension was near ceiling across conditions. In contrast, signs learned in VR were produced more accurately, particularly with respect to handshape. Despite the higher workload, most participants preferred the VR platform for continued learning. These findings characterize key design trade-offs in accessible sign language learning technologies and inform the development of future immersive systems for ASL learning.