AI-Powered Automatic Speech Recognition: Advancing video accessibility on Aalto Panopto
Besides improved accessibility for people who are deaf or hard of hearing captions on videos increased engagement and watch time, as people tend to watch videos longer when they have captions. Captions give a better understanding of the content for non-native speakers or viewers who might have difficulty understanding the spoken language in the video. Also, all videos that teachers produce unless
intended for a time short period of distribution and usage must have captions by Finnish legislation.
Every video uploaded to Aalto Panopto (whether it was created with Panopto or not) is machine-transcribed using ASR (automatic speech recognition technology) and thus making your videos accessible can be achieved with just a couple of clicks. Aalto Panopto site-wide language has been set to English, but individual folders can use a separate ASR language.
To add these machine-generated captions within seconds go to Panopto editing mode and simply select “Captions” from the menu on the left side and choose “Import automatic captions”. You can either edit them to ensure they are close to 100% accurate or simply click“Publish”. Manually editing the captions to be flawless is laborious indeed but the point is that 90% accurate captions in many cases are better than no captions at all.
AI behind speech recognition keeps getting better every day so be curious and have a look at how well the Panopto machine-generation engine handles your lecture now. Panopto AI also handles caption translations into other language subtitles. If you are interested in this new feature, like an example
of having Finnish subtitles of your lecture that is in English, you can join to pilot test this feature by sending an email to firstname.lastname@example.org.