Thread by @LorandBodo

#OSINT tool:
This is a short thread on how to use Microsoft's Video Indexer (VI) to easily extract valuable insights from videos. Analysing open source information, especially hours of video footage is time-consuming so tools like VI can come in pretty handy 1/6

For this exercise, I use a 1h51m long video from a far-right rally in London (Sat 14 Jul), which I downloaded from YouTube & then uploaded onto the VI platform. Uploading time was around 6-8 min, whereas indexing took 44 min (trial account). 2/5

If you take a look at the right side: VI lists all people appearing in the video and (for some) even provides a short bio. It also displays the exact time(s) the person appears in the video (an advantageous feature that could save analysts valuable time) 3/6

Keyword extraction is another great feature that I want to highlight. VI lists keywords that appear in the video (once or multiple times). With just one click, analysts can jump right to the relevant scenes in the video. 4/6

The last thing I want to point out is VI's audio transcription & OCR feature. VI automatically transcribes audio (not perfectly), along with text appearing in the video (e.g. on a banner). All in all, handy features 5/6

In sum, VI is a great tool (free trial) that can help analysts quickly identify relevant content by saving lots of time! If you want to give it a try, go to videoindexer[.]ai and set up a trial account!

Also, huge thanks to @dicooke34 who told me about Video Indexer!

END