Beginning this August, automatic captions will be enabled for videos uploaded to Kaltura.
In the past, we did not enable automatic captions by default because the accuracy of machine-generated captions was often too low to provide a meaningful benefit. Recent advances in artificial intelligence have significantly improved caption quality, making automatic captions a valuable starting point for improving access to video content.
While caption accuracy has improved, automatic captions are not perfect. Content creators remain responsible for reviewing and correcting captions to ensure they accurately reflect the spoken content. This is especially important for videos that include specialized terminology, discipline-specific vocabulary, names, acronyms, formulas, or technical language.
We recognize that captions for STEM, medical, legal, and other specialized subject matter may still contain errors. Automatic captioning should be viewed as a first draft that helps reduce effort, not a substitute for quality review.
To support this transition, you can use this guidance on how to review and edit captions in Kaltura and share best practices for editing captions.
Automatic captions improve access for everyone and reduce the effort required to create captions, but accommodation-related captioning requests will continue to follow the existing CSD process.
If you have questions or need assistance, please contact itaccessibility@uconn.edu.