Thanks to insightface for the great model. It can estimate the 3D shape of a face from a single image.
I will be working on ui+stability of the software as well features such as identify a person and only replacing them in the video, replacing multiple people etc.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
"I have nothing to hide, why should I care about companies collecting my data?" - a thread 📃
First, let's talk about the kind of data that is collected.
Direct: Name, location, face, gender, profession, relationship status, things you have bought or wishlisted, your messages
Deduced: approx income, favorite activities, sleeping schedule, political/religious views, medical issues, academic qualification, if you have pets, types of content you like, screen time, if you have kids, if you moved recently etc.
Now, how do they use this data?
For showing you ads based on your personality, research, training their AIs to deduce more things, influencing masses, selling the data to another company etc.
Now, coming back to the original question. Why should one care? There are at least two good reasons.
cat matches cat
ca+t matches caaaaaaaaaaaat but not ct
ca*t matches caaaaaaaaaaaat and also ct
ca{2,4} matches caat, caaat and caaaat
c(at)+ matches catatatatatat
c(at|orn) matches cat and corn
c[ea] matches cat and cet
c[ea]+ matches caaaat and ceeet
c[A-C0-9] matches cAt, cBt, cCt, c8t etc.
c.t matches cat, c&t, c2t (any char between c and t)
c.+t matches c3%x4t (any number of any chars)
c.*t matches c3%x4t and as well as ct
^ denotes start of a string, $ denotes the end
^a+cat will match aaacat in aaacat but not in bbaaacat
cat$ will match cat in aaacat but not in aaacats
^cat$ will match only and only this string i.e. cat
\d is for digits, \w for alphanumeric chars, \s is for white space chars & line breaks
\D is for non-digits, \W for non-alphamueric chars and \s is for non-white space chars