dChan - Q Origins Project Archive

TYB!

(from this bread's Notables, #22677)

> >>18492348 It would take only 77 anons 1 year to view all of the J6 footage

Which means: it'd take 77 * 365 = 28,105 anons to do it in a day.

Also, Tucker explained there's a lot of "empty filming" like in closets and areas where people normally aren't, or at least, aren't often (e.g., server rooms etc).

Give anons the footage.

We'll figure out how to strip out the extra stuff.

I think Tucker said 41,000 hours of it is "the extra stuff."

So, 44,000 - 41,000 = 3,000 hours.

Even if we have to witness ALL of that, 28k anons (from above) would do it in a couple hours.

Or, those original 77 anons would have it done in about 40 days.

Also we need to use facial recognition software, of which tons exists; here are some open source ones: https://github.com/topics/face-recognition

Most popular, Python command-line-driven:

https://github.com/ageitgey/face_recognition

Anon testing it out now, TYB!

Anonymous ID: 7fe0a9 March 12, 2023, 10:05 a.m. No.18492758 🗄️.is 🔗kun >>2867

>>18492664

anon is retarded

can't do "git clone https://github.com/ageitgey/face_recognition" without it asking for username/password which it didn't used to; oh well! Someone else take the ball and run with it, that's what we do here, kek

>>18492763

got the face_recognition package working, cloned via:

"pip install git+https://github.com/ageitgey/face_recognition_models"

Running is fairly simple, it found the face in "McKek!":

>>18492767

[xyzzy@archlinux ~]$ face_detection Downloads/McKek!.jpg Downloads/McKek!.jpg,178,410,327,261

Then tried picrel2 of 24 faces:

[xyzzy@archlinux ~]$ face_detection Downloads/faces created by AI2.webp Downloads/faces created by AI2.webp,45,419,131,333Downloads/faces created by AI2.webp,342,273,446,169Downloads/faces created by AI2.webp,45,573,131,486Downloads/faces created by AI2.webp,496,266,582,179Downloads/faces created by AI2.webp,192,423,296,319Downloads/faces created by AI2.webp,198,870,285,784Downloads/faces created by AI2.webp,198,266,285,179Downloads/faces created by AI2.webp,45,266,131,179Downloads/faces created by AI2.webp,496,112,582,26Downloads/faces created by AI2.webp,480,883,584,780Downloads/faces created by AI2.webp,45,112,131,26Downloads/faces created by AI2.webp,342,734,446,630Downloads/faces created by AI2.webp,496,573,582,486Downloads/faces created by AI2.webp,496,419,582,333Downloads/faces created by AI2.webp,198,112,285,26Downloads/faces created by AI2.webp,496,717,582,630Downloads/faces created by AI2.webp,31,734,135,630Downloads/faces created by AI2.webp,198,573,285,486Downloads/faces created by AI2.webp,342,112,429,26Downloads/faces created by AI2.webp,352,870,438,784Downloads/faces created by AI2.webp,35,870,121,784Downloads/faces created by AI2.webp,342,573,429,486Downloads/faces created by AI2.webp,352,419,438,333Downloads/faces created by AI2.webp,189,717,275,630[xyzzy@archlinux ~]$ face_detection Downloads/faces created by AI2.webp |wc 24 96 1238

Note that in the above, the second command is "wc" to show # of lines of output, which was 24. There are options to tweak it but this is a great "first look" - this works, and can be used.

The 24 faces image is from this article:

https://www.newscientist.com/article/2308312-fake-faces-created-by-ai-look-more-trustworthy-than-real-people/

>>18492636

TYB!