For videos without audio that can be transcribed (silent videos, videos with only ambient/work sounds, etc.), AI can now read and understand content from the video footage itself.
The extracted content is used for search and AI Chat.
A "Video Content" section has been added to the "Related Information" on the knowledge detail screen.
The extracted content is displayed by timestamp of the scene it originated from.
Clicking on a timestamp allows you to play the video from that scene.
For videos that contain spoken audio, transcriptions and timelines will continue to be created as before.
Related: Checking content read from video footage
https://teachme.jp/8/manuals/51538175
