h a l f b a k e r yWhat was the question again?
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
I recent HB "discussion" that had gotten rather heated
turned from the subject matter at hand to discussing the
apparent mental retardation of the poster. This in turn
led
to the poster countering the claims of his stupidity with
counter claims that the other party was ugly. All parties
eventually
agreed that the opposing party in the
discussion
was both stupid AND ugly.
This is all well and good since discussion from different
points of view are how we all learn about the world
around
us. But to the question of who's the ugliest, most people
in
their right mind wouldn't post a picture of themselves on
an anonymous forum such as this, yet it's a burning
question that begs to be answered.
I propose using the website "Celebslikeme.com" (not
affiliated in any way) or similar website to upload a
picture of yourself and put what your closest celebrity
lookalike is with the percentage match number on your
profile. Mine's some guy named Carmine Giovinazzo,
67%,
(never heard of him) the one where he's in the tank top.
Oddly enough the other pictures of him don't look like
me
at all, just that one. I'm in better shape too. He's kind of
wimpy.
So if you want to make a dartboard of doctorremulac3,
that picture is reasonably close.
Using this or a similar website and posting:
Angela
Joli = 55%, or whatever you came up with would give
people a good enough idea of what you look like while
retaining your anonymity.
Even if you don't post the results, it's kind of
fun anyway.
The site I used.
https://www.celebslike.me/
But I think there are lots of these. [doctorremulac3, Jul 19 2016]
Me (67%)
https://thalifeofli...inegiovinazzo20.jpg
Anybody else? [doctorremulac3, Jul 19 2016]
Flying Toaster (50%?)
http://www.imdb.com/name/nm0343472/
[doctorremulac3, Jul 19 2016]
Loris (48%)
https://www.comedyw...q_profile_thumb.jpg
[doctorremulac3, Jul 19 2016]
Limpnotes (60%)
http://movies.dosth...artini-Pictures.jpg
[doctorremulac3, Jul 19 2016]
BHumphrys (54%)
https://encrypted-t...Ya8lFdyJSq8wPyLktQo
[bhumphrys, Jul 20 2016]
44% RayfordSteele
http://www.imdb.com...63177/?ref_=nv_sr_1
Chris Parnell [RayfordSteele, Jul 20 2016]
Also 43%
https://en.m.wikipedia.org/wiki/Rick_Dees
A face for radio... [RayfordSteele, Jul 20 2016]
59% whatrock
http://adelaiderevi...f-blues-and-laurie/
Hugh Laurie, tho w/o his clever repartee [whatrock, Jul 21 2016]
100% 8th of 7
http://t1.gstatic.c...Gi65GZg2P3flGYaajeA
Uncanny likeness … [8th of 7, Jul 21 2016]
Theircompetitor (Spitting image)
http://popculturebl...IDERMAN_3471540.jpg
[doctorremulac3, Jul 21 2016]
Lurch (match percentage unknown)
http://hairstyles.t...l/Brandon-Routh.jpg
[doctorremulac3, Jul 21 2016]
(??) 46% Loris
http://ia.media-imd...9,0,214,317_AL_.jpg
Jeffrey Katzenberg [Loris, Jul 21 2016]
87% 8th of 7
http://i0.kym-cdn.c...5205.jpg?1303243035
"We're all beautiful inside bro." [Loris, Jul 21 2016]
Wikipedia: Eigenfaces
https://en.wikipedia.org/wiki/Eigenface
In time, there ought to develop a generalised standard library of eigenfaces against which we can compare ourselves in vector form. [zen_tom, Jul 22 2016]
https://en.wikipedia.org/wiki/Eigenface
[hippo, Jul 25 2016]
[link]
|
| |
I like to think I look a bit like Paul Gross, but probably more resemble Adam Sandler having a bad hair day. |
|
| |
looking at that picture of Carmine Giovinazzo... is that the
guy from CSI:something? |
|
| |
My recognition of celebrities is pretty poor, so I'm pretty
chuffed when I recognise someone. |
|
| |
Also I don't think your plan works for people who actually
are
ugly - I scored 48% like Anthony Jeselnik, and to be honest
there isn't much we have in common apart from number of
eyes &c.
I think the site itself knew this, which was why it had
increasingly desperate back-up
suggestions of B.J.Novak(48%), Michael Urie(47%) and Kevin
Pereira(46%)
I'm not posting a link to the photos because if they ever
get wind of this they might sue for
defamation. |
|
| |
...And that's with my best ever photo, from some years
ago.
OK,
so it was a low-res copy - I should try it with the original,
but it's going to be a couple of days. |
|
| |
We're all beautiful inside bro. |
|
| |
Except for the guts and gross parts. |
|
| |
That's actually a sort of interesting idea. Some hotshot
programming wizard could take your face and determine
which famous faces need to be combined in what
percentage to make your face. |
|
| |
So somebody might be 20% Greta Garbo, 30% Laura Bush,
10% Godzilla, 40% Sponge. |
|
| |
This gives me an idea for an art project. As anybody ever
morphed the main players in turning points in history? So
for the 40s you'd have FDR, Stalin, Churchill, Hirohito and
(ugh, I'm getting sick of typing this name on this website)
Hitler. For the 20s you'd have Baby Face Nelson, Al
Capone, Elliot Ness, etc. |
|
| |
I'm a mashup of Jim Carrey, Harry Caray, Mariah
Carey,
Drew Carey, Carrie Underwood, and a carrot. Oh, and
Carrie, the horror film. |
|
| |
First attempt came up Brandon Routh (48%), Christopher Reeves
(43%)... had my hopes up. Then it turned (another picture) to "Pistol
Pete" Maravich (41%, bonus points for the floppy socks) and Sosie
Bacon (41%, penalty points for insulting her with the gender
mismatch). |
|
| |
Well, whatever has your highest match percentage,
that's your "sort-of-yearbook picture". |
|
| |
Those are all ok looking guys. I guess. Not much of a
judge. All guys look kind of ugly to me. Although to
be fair, chicks do cheat with makeup. |
|
| |
Hmmm, "Call for A-list" ... ? |
|
| |
<Looks wave equation and shakes head.> |
|
| |
Not sure about celebslikeme. I just tried it with and without glasses
and with my face at different angles and the matches I get are of
people with or without glasses and faces at the same angle as mine,
so unless there's something distinctive about holding your head at a
23° angle or being seen from that angle, its utility seems rather
limited. Although apparently I look a bit like Richard Ayoade and
Martin Freeman, and I can see that, gender presentation
notwithstanding. |
|
| |
//Hmmm, "Call for A-list" ... ?// |
|
| |
No, the title might be "Halfbakery Member
Approximate Visage Yearbook". |
|
| |
//Not sure about celebslikeme// |
|
| |
Did a little test. I clicked the example guy with the
"Use This Photo" caption and his closest match was
Chris Parnell, the comedian from SNL. I then put
Cirs Parnell's picture into the search engine and it
came up with the exact same picture, but the
number was only 80%. Hmm. But then I tried Kathy
Bates the actress and they came up with a
different picture of her rating the match 100%. |
|
| |
I've been told I'm the spitting image of Doc Ock (or Al
Molina) |
|
| |
I've just tried it with my original higher-res mugshot, and it sits there, twirling away to itself and doesn't resolve.
I think I broke it. |
|
| |
[edit] OK I finally got it to work. Apparently I look a bit (46%) like Jeffrey Katzenberg, CEO of DreamWorks Animation. He has less hair but whiter teeth than me, but it's a better match I think. |
|
| |
Loris, I got a "Referral Denied" when I clicked the
Katzenberg link. |
|
| |
Here is how such endeavors can take place objectively and scientifically. |
|
| |
1. Google image your name. For example doctor-remulac returns an assortment of images that probably look more or less like doctor_remulac. |
|
| |
2. Use an image merge function to merge the faces, obtaining the average. |
|
| |
I think a custom merge function would be good. Image recognition sorting out the mug shot type faces you want from faces in profiles, pictures of sea cucumbers, powerpoint graphs and so on. Also current merge programs want 2 faces; this needs to be customized to use many. |
|
| |
// current merge programs want 2 faces // |
|
| |
Hilary Clinton could use one all on her own, then. |
|
| |
Question; using a merge programme for faces, would (A + B) + C look the same as (A + C) + B, &&|| A + (B + C) ? |
|
| |
It's not simple arithmetic addition, but it should be easy to test. |
|
| |
My cat said, that I look like a budgie. |
|
| |
If you're just averaging points, and the first two are a separate operation to the next, then the third face would have a higher weighting. |
|
| |
Yes, but that depends on how the programme weights the first two inputs. If it uses centroiding rather than simple averaging - which would be reasonable - then the bias would converge on the median. |
|
| |
Rather disappointingly, I turned out to look exactly like me. |
|
| |
//the third face would have a higher weighting// |
|
| |
Presumably you'd want to start with a power-of-two
cardinality, then pair them up as if they were competing at
tennis. |
|
| |
This is not completely dissimilar to the 'Eigenface'
approach to automated facial recognition (see link),
but using celebrity faces as the eigenfaces. |
|
| |
Oh c'mon! I look nothing like James Lipton, but if it helps me fill my celebrity autograph book I might be able to fake it for a few episodes. |
|
| |