Singing Without a Larynx: A Vowel-Based Real-Time Interface for Embodied Vocal Performance

This demonstration presents a web-based interactive system designed to support vocal expression for individuals who have undergone a laryngectomy. Using computer vision and vowel classification based on mouth shape detection, the interface enables users to generate real-time vocal synthesis without phonation. The system aims to expand conventional understandings of voice beyond speech recovery, offering a space for sonic and musical experimentation grounded in embodied interaction. Through the interactive app we aim to explore how participatory design and critical disability studies can inform the development of assistive vocal technologies that move beyond compensatory or rehabilitative paradigms. Rather than attempting to restore normative vocal function, the system invites a reimagining of voice as spatial, malleable, and co-constructed through interface, gesture, and intention. Anchored in a critique of ableist assumptions within technological design, the project aligns with identity recovery, creative autonomy, and user agency for individuals who have experienced structural silencing. Attendees will observe the vowel classifier in action, experiment with input-output relationships, and engage in discussion around the system’s conceptual and technical foundations. Fully browser-based, the interface integrates flexibly into live performance or research settings. By centering aural diversity and embodied musical interaction, this demo contributes to the CMMR 2025 theme “Sound, Music: Space, Place” through a redefinition of voice as a site of expression that can be felt, shaped, and reclaimed.