Beginning later in April 2018, Video on Demand (VoD) will begin automatically generating captions to all newly uploaded videos, using Automatic Speech Recognition (ASR). Previously, the service only added captions when video owners specifically requested them.

The service will also begin retroactively adding captions to previously uploaded videos that do not already have them. This moves the service further toward meeting the accessibility standards outlined in the university’s strategic plan.

Because ASR-generated captions typically only provide 70-80% accuracy, video owners should review and edit the ASR-generated captions to improve accuracy. This should be done both for newly uploaded videos and, over the coming weeks, for previously uploaded videos to which captions have been added. Video owners should review their automatically generated captions with an eye toward both textual errors and factual accuracy that might impact usability.

The VoD service already provides tools for editing captions in previously uploaded videos. For more information, please see the article Edit Captions in Video on Demand.

For general details about online accessibility at Cornell, please visit the Web Accessibility page.