Preprint
EvolveCaptions: Empowering DHH Users Through Real-Time Collaborative Captioning
Liang-Yuan Wu and Dhruv Jain, arXiv (preprint), 2025
Read the originalRead on October 16, 2026
- What it is
- A 2025 study from the University of Michigan of EvolveCaptions, a research system in which hearing participants correct live speech recognition errors and the recognizer is fine-tuned to the deaf or hard of hearing speaker, tested with 12 deaf and hard of hearing participants.
- How much weight it carries
- Preprint from a university research group. The sample is small, the system is a research prototype rather than a product, and the abstract reports the direction of the error reduction without the size of each change.
Facts taken from this source
- In a 2025 study of EvolveCaptions, a research system, word error rate fell for all deaf and hard of hearing users within one hour of use, with about five minutes of recording time on average.
“EvolveCaptions reduced Word Error Rate (WER) across all DHH users within one hour of use, using only five minutes of recording time on average.” (Abstract)