Real-Time Caption Challenge: C-Print
Michael S. Stinson, Pamela Francis, Lisa B. Elliot, Donna Easton · 2014 · Proceedings of the 16th International ACM SIGACCESS Conference on Computers & Accessibility (ASSETS '14) · doi:10.1145/2661334.2661337
Summary
This demonstration paper presents C-Print, a typing-based real-time captioning system developed over 25 years by researchers at the National Technical Institute for the Deaf (NTID) at Rochester Institute of Technology. C-Print provides communication access for deaf and hard of hearing (D/HH) individuals in educational and professional settings by having a trained transcriptionist produce a near-real-time text display of spoken information. The system was showcased as part of a real-time caption challenge at the ASSETS 2014 conference. Unlike stenographic captioning (CART), C-Print uses a standard keyboard with a computerised word-abbreviation system based on phonetic rules, allowing the captionist to type condensed abbreviations that automatically expand into full text. For example, typing "wlkm to t nxl tknkl nsttt f t deaf" expands to "Welcome to the National Technical Institute for the Deaf." The spoken content appears on a computer or mobile device approximately two seconds after being spoken, and can be saved and distributed as a transcript afterward.
Key findings
The paper highlights C-Print Mobile, a then-recent development funded by the National Science Foundation that extended the system to mobile devices. C-Print addresses a key gap in communication access: different D/HH individuals have vastly different needs — some use sign language interpreters, others rely on speech-reading and hearing aids or cochlear implants, and many prefer text-based access. The captioning approach recognises that many D/HH people are strong text consumers who already read captions on television and use text messaging extensively. C-Print's abbreviation system enables trained captionists to achieve the typing speed necessary for real-time transcription without requiring the specialised stenographic keyboard and extensive training period associated with CART services. The system produces both a live display and a post-session transcript, giving D/HH users both real-time access and a reference document.
Relevance
This paper is relevant to accessibility practitioners working on communication access in education, workplace, and conference settings. It illustrates that real-time captioning is not a single technology but a spectrum of approaches with different trade-offs in cost, training requirements, accuracy, and verbatim fidelity. For organisations choosing between captioning services, C-Print represents a middle ground — more accessible to train for than CART stenography, while still requiring human expertise rather than relying solely on automatic speech recognition, which in 2014 was far less accurate than it is today. The mobile extension (C-Print Mobile) foreshadowed the broader shift toward mobile-first access services. The work remains relevant as a benchmark for comparing human-driven captioning against modern ASR-based alternatives.
Tags: deaf and hard of hearing · real-time captioning · communication access · transcription · mobile accessibility · education