Computer vision basically isartificial intelligence that focuses on imagery. It's the same...
Join the conversation
Log in or create an account to leave a comment
Log In
Computer vision basically is
artificial intelligence that focuses on imagery. It's the same kind of technology
that's used with autonomous vehicles, and it's used to compare
the different frames and to create the end conform. If you look at this image right here, there are a number
of different interest points, and it doesn't matter
that the reference cut was cropped. Those interest points are
the same between these two things, and that's what the algorithm does. It's able to match those interest point,
and then basically determine a confidence that,
"Yes, this is the same frame." From a technical standpoint, we haven't worked
with this much footage before
artificial intelligence that focuses on imagery. It's the same kind of technology
that's used with autonomous vehicles, and it's used to compare
the different frames and to create the end conform. If you look at this image right here, there are a number
of different interest points, and it doesn't matter
that the reference cut was cropped. Those interest points are
the same between these two things, and that's what the algorithm does. It's able to match those interest point,
and then basically determine a confidence that,
"Yes, this is the same frame." From a technical standpoint, we haven't worked
with this much footage before
Full Transcript
00:00:01.000 --> 00:00:04.254
Computer vision basically is
artificial intelligence
00:00:04.337 --> 00:00:05.839
that focuses on imagery.
00:00:06.256 --> 00:00:09.759
It's the same kind of technology
that's used with autonomous vehicles,
00:00:09.843 --> 00:00:12.762
and it's used to compare
the different frames
00:00:12.846 --> 00:00:14.806
and to create the end conform.
00:00:16.001 --> 00:00:17.892
If you look at this image right here,
00:00:18.999 --> 00:00:20.052
there are a number
of different interest points,
00:00:20.603 --> 00:00:23.069
and it doesn't matter
that the reference cut was cropped.
00:00:23.773 --> 00:00:26.484
Those interest points are
the same between these two things,
00:00:26.568 --> 00:00:28.528
and that's what the algorithm does.
00:00:28.611 --> 00:00:31.698
It's able to match those interest point,
and then basically determine
00:00:31.781 --> 00:00:34.409
a confidence that,
"Yes, this is the same frame."
00:00:35.285 --> 00:00:36.995
From a technical standpoint,
00:00:37.012 --> 00:00:41.999
we haven't worked
with this much footage before.
Want This Clip in HD?
Upgrade for HD/4K downloads and unlimited access. Upgrade now →
Movie Summary
A glimpse behind the scenes into the complicated process of recovering and completing the final film of legendary director Orson Welles.
