Does auto caption learn?

Contributor

We are noticing common mispellings of proper names and are updating the transscripts. Is there machine learning on the backend that will eventually improve our tenant transcrption so it makes better guesses at internal terms?  

1 Reply
I recently watched the below video about Stream features and back-end design and I'm pretty sure they covered machine learning (ML) when correcting automatically generated captions. I haven't seen this UI in practice though but I've only been briefly trialling Stream internally.
Maybe I simply missed it, maybe it was an as-yet unreleased feature, maybe I require an E5/Stream Plan 2 licence to view it.

https://stream.microsoft.com/en-us/blog/learn-how-microsoft-stream-was-built-using-azure/