Advice on automated captions
Currently our video content, delivered via Kaltura MediaSpace, uses closed captions generated using automated speech recognition. This is about 70-85% (Kaltura, 2018) accurate depending on the topic, audio quality and lecturer.
We advise you not to rely solely on the captions when reviewing your video content. These are considered a supplementary resource. Make sure you check the captions against the PowerPoint and readings to ensure you have a good understanding of the topic and the terminology used. You are encouraged to seek clarification from your teaching staff if necessary.
Our closed captions appear as a text description at the bottom of the video which you can switch on and off.
If you require accurate captions as part of your reasonable adjustment, contact email@example.com to discuss your requirements for quality captions. This will then result in human intervention to make corrections on the automated captions. This usually substantially improves the accuracy of the captions, but the accuracy can still be impacted by terminology used, accent of the lecturer and audio quality of the recording.