Something good that has come from this global pandemic and staying home is the new and renewed interest in online captions and transcriptions. Here's a new article from the MIT Technology Review describing the challenge and some solutions.
The pandemic made life harder for deaf people. The solutions could benefit everyone
We tested Otter.ai, which records/transcribes live conversations using AI, during an online training for which live captioning was actually provided. The Otter.ai tool provided us with captions, on our phone, faster and they were more accurate than the captions provided on the computer. We were also able to save the captions/transcription, instead of only being able to see the captions in real time.
For video captioning this spring, we had one of Lily's teachers download all his videos into YouTube and send her a private link. This way the videos were automatically captioned, without too much effort. The school didn't provide captions on their VR video platform, although it looks like they could pay for them. Lily can hear the online videos pretty well, but she definitely prefers videos with captions, especially if you can't see the teacher's face.
Lily also used Zoom for many of the office hours her teachers held. She could follow the conversations well if the teacher was talking, but it was much more difficult if it was a group chat, just like in the classroom. This is where captions could have been a big help. Looks like Zoom is working on incorporating a caption tool for the end of summer (currently there is a third party work around), but Google Meet, Skype, and Microsoft Teams already have this feature. Others rave about AVA's integration, but we haven't tried it.
http://connect-hear.com. This site is a terrific Knowledge Base of resources for people who are deaf or hard of hearing. It includes things like mask resources, video conferencing options, captioning, and assistive technology. Definitely a site to investigate.
Finally, in another life, we produce lots of podcasts. We use a tool called Descript to create podcast transcripts. Everything we produce now has captions and it's quite a bit of work to make sure the captions are accurate. However Descript gets us to about 75-80% of the way there.
Comments