ThePatr10t
commented
3 years ago Audio to text transcription is a helluva good idea. Noted.
Right now Im expanding my efforts to find and download videos with a wider radius from the capitol building, however my efforts in getting them up to the site have been slowed. Youtube has lots of bandwidth for the capitol videos, but they need to be uploaded one-by-one. Im investigating other ways of hosting the many other videos with batch-uploads, but this is taking time.
At the very least, the facial recognition stuff Im doing now will be on a much larger pile of videos than that exposed on the map.
I noticed that all of the points on the map are only at the Capitol. There should be at least some juicy (i.e. incriminating) takes that were uploaded before and after the attack.
Sorting the parler video uploads by date, collect all uploads with metadata corresponding with the dates just before as well as days after the 6th, that are also in the greater DC downtown area and put them on the map as an alternative color points.
Has there been any effort to make available auto generated English language transcript .srt or .txt filesof the videos? Having a greppable dataset would streamline the shit out if identifying any juicy uploads that may not be at the actual capitol itself during the attack.