In the actual photo? Is there any meta data? Is there any EXIF data? What was the photo filename, and what does searching for that (with increments up and down) return?
To try and guess what happened:-
1) One of them followed you and got information. (You say this didn't happen.)
2) They had information about the event and worked from there.
3) There were other people in the photo, and they got face-matched, and their friends lists were not hidden, and they got it from that.
4) They had snippets of other information and just trawled through very many searches to get the name.
I'm a bit confused how they didn't already have her name. You told them you met someone, but didn't give any kind of name?
If they knew her high school and were doing lots of batch processing, I wonder if they found digital copies of the yearbook and tried facial recognition. For a single subject, it doesn't even need to be particularly accurate- they can easily sift through twenty possible matches with their own eyes.
I'm sure there's a lot of information missing from this story, which leads to a lot of possible solutions that aren't negated by the lack of limiting circumstances.
More than likely the author is an idiot and the image had EXIF data. Considering they already knew her highschool, it wouldn't had been to hard to go through the year books in a few hours.
Heh, geek blog effect. Reddit et al have created some very bad habits of dialogue, such as assuming the subject of a link isn't reading it. And being nasty with candor.
I still remember the time I said something impolite about David Brin. Brin took offense and replied.
I'm actually a huge fan of David Brin. That sucked.
I'm not sure where you go the idea that pseudo-celebs and IT managers are immune to stupidity or posting stupid articles. There's always an element of "who cares if my content sucks, lets just get it done" when posting bloggy stuff, and it's pretty apparent here.
Calling EXIF "back end stuff" is pretty stupid, I don't care if the author's ego is bigger than that statement, it's still true. The article might as well be how many apples fit in a bag or how long is a string.
@asymetric - I am not an engineer, I don't know how a lot of these things work. Rather than critique, I would appreciate a better analogy that I can use in the future
HN tries to avoid that kind of rudeness, but it happens. Like I said, geeks can get bad habits from the locker-room culture online. You'll notice that your critic's posts are gradually fading to grey thanks to all the downvotes he's accumulating (established users get a downvote button).
The same two engineers who found out who she was? :-)
Seriously: given your story, if, say, they are colleagues, too, they may be in on it, and lie to you about the EXIF data (if they _are_ engineers, they certainly lied; how can there ever be 'nothing interesting' in EXIF data :-)?)
Most of the time, the simple answer is the likely one. My money is on the high school-yearbook line, but it might have been anything (did one of your colleagues go to the same school?).
A photo without EXIF data? Someone must be trying to hide something from me. Let's see how good he is, and check whether there still are parts of the file with metadata in the disk's free blocks.
"or replaces it with metadata from another photo"
A Nikon F5 at f/5.6 and 1/100s? No way! If so, that car must have done a thousand kilometers an hour or so. Also, the aberration looks more like that of an Canon lens, but I'm not 100% sure of the model. I wonder whether it is possible to train a model on the 'JPG to camera model' problem.
"or makes up some bogus metadata."
Hm, I thought the Eiffel Tower clone in Japan was in Shanghai. The GPS coordinates seem to indicate that Tokio has one, too. Let's google to check that.
OK, that's more for the hacker engineer, but I thought one would not have to make that explicit on HN.
>Someone must be trying to hide something from me.
Umm, No, many popular photo-sharing sites at least give you the option to strip metadata from photos when posting. IIRC, fb does this as well, by default.
> and check whether there still are parts of the file with metadata in the disk's free blocks.
Good luck.
>A Nikon F5 at f/5.6 and 1/100s? No way!
It doesn't have to be believable metadata to be practically useless.
>Also, the aberration looks more like that of an Canon lens, but I'm not 100% sure of the model.
> I wonder whether it is possible to train a model on the 'JPG to camera model' problem.
Sure, but not for beer money. You could probably even have some success identifying individual cameras by characterizing their CCD/image sensor, and comparing to known photos.
The fact that you required two engineers to do something that you could have confirmed yourself with 2 seconds of Googling shows that your ego is a bit inflated.
I think the key is knowing what high school she went to. From there, you could possibly get yearbook pictures and student lists. Knowing an approximate age would narrow the possibilities significantly as well.
I would probably take a social engineering approach to this problem- if you knew her high school and approximate age you could probably find someone who was in that class who would know her pretty easily. At that point it's just a matter of either tricking them into identifying the photo or just straight-up asking them.
Assuming that she didn't show up in a reverse Google Image Search, maybe they used some sort of facial recognition API (like http://www.lambdal.com/), perhaps along with the other people in the picture, against scraped pictures from LinkedIn and Facebook based on her appx age, location (Austin), gender (F), and high school.
At this point he could not see my screen (he was standing directly in
front of me), but offered me a simple bet – He said if I sent him the
picture of the girl he could find out her name.
Reminds me of:
Sky Masterson: One of these days in your travels, a guy is going to show you a
brand-new deck of cards on which the seal is not yet broken. Then this guy
is going to offer to bet you that he can make the jack of spades jump out of
this brand-new deck of cards and squirt cider in your ear. But, son, do not
accept this bet, because as sure as you stand there, you're going to wind
up with an ear full of cider.
> Simplest explanation: Someone saw you and told him.
Of course one of the first things I did was ask another friend who I thought met her. But no, ultimately the answer did come from the internet, not from anyone I knew.
I'm the engineer he's referring to. I still get a laugh seeing Justin's amazement when we told him her name. To answer a few questions...
- This isn't the actual photo but looks similar, except that she was alone. Justin was afraid of reposting the same photo after the results of handing over a photo last time :)
- There was no useful exif data in the image.
- We knew her high school and approximate age in Austin, but were unable to find any yearbooks / class rosters online. (Idea not pursued: try to acquire a yearbook offline).
- Facebook Graph Search and LinkedIn with the criteria we knew didn't help.
- Google Image search didn't help.
> "I'm sure there's a lot of information missing from this story"
Quite true.
Ultimately we had several methods going at once to try to figure it out.
The one that came back with results first relies on some information Justin still doesn't know we had (though he would still be impressed), but other approaches we also expected to work would be possible with only what's in the story, and haven't yet been mentioned here.
Justin's own phone leaked his location (through some google-lattitude, or Yelp like service or perhaps gps directly) while Justin drove (or rode with her in a taxi) to her home.
How would they have gotten my GPS location? And how would they have figured out which one was her house. I was in her neighborhood and stopped multiple places that were 'significant' to her as a kid. Any of them could have been her house. And I wasn't at anyone for more than 60 seconds
You obviously knew something about her that made it possible, which you even admit Justin doesn't know you had. People are getting bored with your shenanigans, spill it!
I don't feel bad about letting the discussion continue as-is because of the fact that we completely expected to get the answer with only the information mentioned here. I still haven't heard this method mentioned yet directly.
The "extra" information we got came later after we already started working on it, and came as surprise. It just happened to bring the result back sooner than the other approach.
Nah I like this approach...let us bumble around for awhile in the discussion...I'd be very surprised if no one on HN could surmise the solution, but I'll admit I'm stumped (though I've only thought about it for a few minutes before getting back to work)
So you stumbled on a piece of information which proved useful, notwithstanding that you didn't intend to look for.
If Justin observed correctly, you guys were very impressed with yourselves. Had you retrieved the answer by methods of which you were confident, you would've still been impressed, but not greatly.(I'm making a lapse conjecture here, of course)
Since you were confident that you would be able to find her, prior to looking at the picture, according to the article, I don't think you were looking for any information in the picture other than for confirmation.
However, some extra information in the picture(since it was taken very recently) enabled you to look further, which returned a result. Perhaps a building, an address, or some sort of landmark-ish location which narrowed whereabouts of the picture vastly? If this were the case, then I can understand that you guys were very impressed, because well, it is.
Can you or did you mention what it was that Justin didn't know you knew about her. Seems rather unfair, to him and us.
So we could all plow through every event that happened in Austin that she may have attended, but without knowing what part of town it starts looking rather boring. I have a real job. :)
So it seems you weren't even confident when you got your confirmation, a.k.a., it was an informed guess but could have just as easily been a girl that looked alike at an event at the time she was supposedly at an event?
* Everyone was at the original bar at the same time - someone else had photos from the night that had her and other people in them, and those people (her friends) could be found online.
* The bar had some kind of "sign-up for x" list, or some kind of contest, or some kind of photos of the night, which were helpful.
* You searched twitter/facebook/whatever for a girl posting about a boy she just met at bar X, blah blah
* We know she was going to some kind of event - you guys figured out what that was, and found a guest list or something.
* You guys did a some wi-fi sniffing and read a few emails. He did say he emailed you the photo, so you know which account to listen for.
So in the other post, you said you inadverdently realized you had a second photo of the target (which presumably contained key identifying info)...so, are you saying here that that second photo didn't come from investigating the event that she was at (i.e., if it had a meetup/lanyrd page, looking at all the attendee photos and seeing what looked like her?)
The second photo that we later realized we had was one a teammate had taken himself and didn't have anyone else in it. I think it was doable with just 1 photo, and I'm not ready to share with Justin how the 2nd photo helped. :)
In a single picture where she was a random bystander? Unlikely that you'd be able to read anything that had her name on it, unless it was an actual name tag. Even then, people don't generally stand very still for photos that are taken of other people.
An article of clothing or accessory of some kind from a conference she attended that helped you look up attendees and identify her, or some other identifying club/service that contains profile pictures?
If it were me, the first thing I would do knowing her high school is create a Facebook account and upload her photo (with her face) and see who Facebook suggests I tag in it.
I'm not sure what the rules are concerning tagging suggestions, but I have noticed that it suggests that I tag people I know in other people's photos so you KNOW they recognize everybody in every photo - the trick is meeting the right conditions for it to allow it to share that recognition match with you based on her privacy settings I think...
The team(including Justin: using first name here for convinience's sake) had a trip to Austin. Where they were should be a public knowledge. Perhaps they(you) started the search from there. A facebook search or so could tell some plans of people who would have fun at that particular place.
It is also possible that during the few days' correspondence one of the engineer guys had some other piece of info that we are not familiar with. Seeing how the engineers already knew her high school, I'm sure Justin talked about her with co-workers.
I think the key information that's missing here is what the engineers knew but Justin didn't know that they knew.(Sounds like a quote out of a Friends episode, mmm)
You looked on Instagram and other for pictures taken at the bar, and you found one either taken by the woman in question or one of her friends with the mystery woman tagged.
Or maybe you knew that she wanted him to meet her at an event and you found a roster for that event- or ie if it was a real estate convention you would have a pretty good idea that she was a realtor and could look at websites for realtors in Austin.
Or maybe you checked his recent facebook friends, looked at his recent twitter followers, and everyone who liked/commented on his recent photos/posts on Instagram and whatever other social networks he's on- as well as the photos/posts he liked/commented on.
I think you took a (relatively) low-tech approach to this.
post pic on craiglist/similar in Austin area. Say you found wallet/purse/camera/etc with her picture. Anyone that knows her please call/email w/ Name/Address for you to send back. Use highschool for validation when contacted. Might have some error.
This is the closest I've heard to one of the ideas that we expected to work, so I'll share.
- Create an ad on Facebook with her picture.
- Target people that went to her high school within 4 years of the range of years we thought she graduated within. Was only a few hundred people so the cost of running the ad would likely only be a few bucks.
- The ad copy / landing page just needed to be convincing that we weren't stalking her. We went with the "help us win a bet" approach but the "lost camera" approach would have been good also.
Ultimately we got the answer sooner from another approach after we found out that we had another picture of the girl that Justin didn't know we had. So we canceled the ad, but I think this would have worked if given enough time.
yah I was actually thinking of facebook ad just after I replied, with that criteria..only thing that turned me off from it what that you made it sound as highschool was a post-processor (validation) vs pre-processor (targeting)
This method likely would have worked by itself. The method that worked sooner wasn't just "another photo" found online of her friends as grecy mentioned. It was a photo we had taken ourselves, and it didn't have anyone else in it. But I'm not ready to tell Justin more details about it.
This strikes me as odd, albeit a clue. You had a photo of a girl you'd presumably never met. A photo you'd taken, that you didn't know was her at first.
They went to the same event, took some photos. And in checking their photos, see that the same girl was in their photos too. Then they saw who she was with, and were able to locate a friend of hers. From locating a friend of hers from this second photo, they located her.
Perhaps she was doing a presentation, so was alone at the front of the room. The photographer, being in the audience, would obviously know which presentation it was and from there its a quick look-up on the list of presenters.
You had another picture that showed where she worked? Used LinkedIn and company's webpage (Some company list their employee) to narrow it down to a few people. Or send a message to a coworker on linkedin whatever looking for information about that girl and they gave it to you? I'm curious :P
sophistry [ˈsɒfɪstrɪ]
n pl -ries
1. (Philosophy)
a. a method of argument that is seemingly plausible though actually invalid and misleading
b. the art of using such arguments
2. subtle but unsound or fallacious reasoning
3. an instance of this; sophism
Your reply appears to accuse the parent of sophistry, but you provide an even worse argument for your claim than the parent does. It is not obvious to me how posting the definition of “stalk”, according to which the bet was not stalking, is an invalid or misleading argument.
I am directly accusing the parent of sophistry by selecting to post a definition of "stalking" that required malicious intent if not violence when that is not the totality of the definition, is not consistent with Merriam-Webster or Oxford dictionary and seeks to excuse their behavior by the omission. .
Taking the third definition of off merriam-webster online doesn't really prove your point here. There is a reason why the title of the op was "Why Engineers Scare Me". I know this was all in good fun but please do understand why this is stalking and how the same technique can be used in very harmful ways.
It was this line -
"On Thursday night, after exchanging texts all day, around 8:00pm local time the girl sent me a picture of her in order to convince me to meet her at an event she was going to."
Find the event Thursday night at 8:00pm in Austin and start tracking down from there.
...which, possibly, would include a picture of that person either in the same outfit or include a profile picture of that person commenting or liking the event.
Your facebook link contains a list of 73 persons going to the event, so maybe one of them had a picture that matched the one the team had of the girl...
Philfreo > I think that the girl did share something with you that Justin doesn't know, such as something that she did accomplish in that school, like winning some kind of award, winning a competition or something similar. Then, you used that information along with the name of the school in Google.
So you know the school and you have her picture. By her picture you can make a guess about her age. Then you search the web for graduates for particular dates from this particular school. It doesn’t have to be a yearbook or a class roster, people on LinkedIn mention the schools they’ve attended. Sooner or later you’ll find someone and then you can narrow the search. Even simpler, once you find someone you simply contact him with a fictional story as to why you’re searching this woman. From there on it just takes some old good fashioned social engineering.
Wherever you were that night had a Facebook page with an event created for the evening you were there. You uploaded your second picture to the Facebook event, or tagged the Facebook page of the bar you were at. Then someone tagged the girl in the photo, revealing her identity.
The second photo you realized you had showed the subject holding the device used to take the picture sent to your colleague. You investigated that particular device and found a way to uniquely identify it's pictures based off some characteristic of it's photos, not necessarily EXIF, then scoured the Internet for photos with the same unique characteristic until you found one that revealed more information about the subject?
If your engineers as self-satisfied about finding her as it sounds, then they'll be happy to describe their process over a beer or two. I've never known an engineer who doesn't like discussing how they solved some riddle.
The engineers were made to beg for info on the girl, so they might be getting some satisfaction from keeping a secret from the author in return. I know I would. Give 'em a taste of his own medicine. Also, if it was a relatively simple solution, it's probably more exciting with some mystery overlaying it rather than just explaining it outright.
How are you surprised that they did it when they knew what high school she went to? Two possibilities:
Option 1: Step 1) Find copies of yearbooks going back a few years, 2) find her face in the yearbook.
Option 2: Step 1) Search facebook for people who went to the same high school as she did in the probable time frame. 2) Look for her in that set, or in the set of friends that the people who matched the search have.
Feel free to sprinkle in face detection on the yearbook/facebook photos to make it more engineer-ey.
As a matter of fact, how did they find out what high school she went to? I can't imagine you telling them that if you were not providing much information about her to anyone else.
They knew what high school because it was a slip up. If you knew these guys, they aren't the kind who would just sit there and look at pictures. They had some software running.
> If you knew these guys, they aren't the kind who would just sit there and look at pictures.
As someone who has been a software developer for quite a long time, this statement makes a big assumption.
A good software developer knows it is often far easier and faster to flip through a couple of hundred photos and process them with your meat-brain rather than write a one-off bit of software to do the matching, even if you start from an existing base like OpenCV or whatever.
Having software running doesn't preclude using your brain - first to consider possible heuristics and then to do the types of analysis computers aren't good at. There's a reason mechanical turk exists.
Where was the photo taken? What was in the photo other than the girl?
If the photo was taken at her home and there is any landscape in the background, the obvious initial point of attack would be to identify the location the photograph was taken.
If you know the high school and year(s), you go thru the friends list of every person on Facebook who did post their high school from that year (or nearby years), and does show their friends, until you see her face.
That's only a couple hundred kids, each with a couple hundred friends.
If you still get nothing, start doing friend requests for the people that don't show their friends. I'll bet you'd get another 25-30 kids that way.
I don't see why you'd need facial recognition software to go thru 20,000 pictures.
3/4ths of them would be the wrong gender/age and be immediately disqualified.
But if you're a programmer, you can probably scrape and aggregate to speed things up - drop out the males, other high schools, only show pictures once, etc.
Philfreo > I think that the girl did share something with you that Justin doesn't know, such as something that she did accomplish in that school, like winning some kind of award, winning a competition or something similar. Then, you used that information along with the name of the school in Google.
I wonder what the resolution of the original image was?
The inside of her thumb is visible through the glass, so with enough detail it might be possible to record her thumb print. Even if you could do that, though, odds are against her print being present in any publicly-available database.
One of the things that looks like a clue to me is that this the lead engineer of close.io. Browsing through their site, it looks like they do "telecom in an API"; it reminds me of Twilio and OneBox.
So, here's a thought. Let's say the Engineer has access to the SMS logs (NSA-style) for Justin's phone. (Maybe his phone goes through their system. But even if not, maybe it's company-provided, so there's an admin interface provided by the telecom that lists this information.)
Justin probably doesn't regularly text people in Texas, since the team is based out of Palo Alto. So look through his logs for a number in a Texas area code (http://en.wikipedia.org/wiki/List_of_Texas_area_codes) that he's been texting more often than usual.
That gets you the girl's phone number; probably a cell phone. Now, you need to go from the cell phone number to a last name. Reverse Phone Lookup probably won't work for this, although it's certainly worth a first try. More likely candidates: looking through the users table at close.io for a matching phone number, or doing a google search for the phone number. Or, try calling in the middle of the night and seeing what the voicemail says. Any of these approaches might work.
Once you've got a last name, grab a copy of the white pages. Most public schools publish their district boundaries, so go grab a copy of that. Look through all entries that match the last name, and see how many of them fall within (or close to, in the case politics made the boundary change) the district boundary. You probably will get a handful (maybe three or four). If you only get one, bingo!
If you get more than one, I'd call the cell phone to see if I can get a first name somehow. Then, I'd go through the landlines in the White Pages and call them one at a time, "Hello, may I speak to X? Sorry, wrong number." until one matches.
I love this explanation. But unfortunately there was no telephony hacking of any kind as this was just an ordinary personal cell phone.
I did have an idea that we could try to hack his Verizon account to look at his SMS log, figuring that Forgot Password questions might be easy or that we could intercept an email. But we didn't do this.
Since the point of this article is probably to promote their company, having the answer be "our engineers looked at your private information for a prank" seems like a no go.
No need to call in the middle of the night to see what voicemail says. Use slydial.com and it connects directly to voicemail without ringing. If you hang up promptly, they get no record of a voicemail being left. I do this for most unknown numbers which call me.
In additional to using TrueCaller.
Saying she "doesn't exist on the internet" is a big assumption. First, she is on Facebook, so she's on the internet, easily accessible or no. Second, most people leave some internet trail through online forums, newspaper articles and newsletters.
I find it very likely that she was found on the internet somewhere.
So basically the engineers knew something specific about the girl already which narrowed their data set. Unless they tell you what that was, we can go all day long guessing how they "figured" it out. In fact with this information I am interested in the actual time taken to find the girl. They could have found her in five minutes after receiving the picture and you wouldn't know. Why? From the article:
1) They were running software on a computer for hours
They are engineers!
2) The first couple of tries didn’t work
Are you sure they weren't debugging?
3) They knew what high school she went too prior to searching
If they know the high school they only have to find one or two established teachers who were around for a long period of time. It is not unlikely they can call a name from a picture after all those years.
This smells like viral marketing to me (too bad I still don't know what close.io is, should have added some information to the article. Or are you looking for new engineers? "Hey look! At close.io we seal bets with engineers. About girls! Come work for us!").
He said she only bought the dress recently. They did a image search and found a local store that stocks that particular dress - and she had checked in with foursquare?
219 comments
[ 3.2 ms ] story [ 241 ms ] threadTo try and guess what happened:-
1) One of them followed you and got information. (You say this didn't happen.)
2) They had information about the event and worked from there.
3) There were other people in the photo, and they got face-matched, and their friends lists were not hidden, and they got it from that.
4) They had snippets of other information and just trawled through very many searches to get the name.
I'm a bit confused how they didn't already have her name. You told them you met someone, but didn't give any kind of name?
I am positive they didn't have number 2
There were no other people in the photo
What other snippets of information would be useful?
And they never met her, only I did. So they didn't know her name. And yeah, I said "I met a cute girl." I didn't say I met "name"
More than likely the author is an idiot and the image had EXIF data. Considering they already knew her highschool, it wouldn't had been to hard to go through the year books in a few hours.
I still remember the time I said something impolite about David Brin. Brin took offense and replied.
I'm actually a huge fan of David Brin. That sucked.
Calling EXIF "back end stuff" is pretty stupid, I don't care if the author's ego is bigger than that statement, it's still true. The article might as well be how many apples fit in a bag or how long is a string.
edit: I'm aware of the irony in criticizing online geek culture and then bringing out a foul-mouthed meme pic.
Seriously: given your story, if, say, they are colleagues, too, they may be in on it, and lie to you about the EXIF data (if they _are_ engineers, they certainly lied; how can there ever be 'nothing interesting' in EXIF data :-)?)
Most of the time, the simple answer is the likely one. My money is on the high school-yearbook line, but it might have been anything (did one of your colleagues go to the same school?).
If one strips all of the metadata a photo, or replaces it with metadata from another photo, or makes up some bogus metadata.
A photo without EXIF data? Someone must be trying to hide something from me. Let's see how good he is, and check whether there still are parts of the file with metadata in the disk's free blocks.
"or replaces it with metadata from another photo"
A Nikon F5 at f/5.6 and 1/100s? No way! If so, that car must have done a thousand kilometers an hour or so. Also, the aberration looks more like that of an Canon lens, but I'm not 100% sure of the model. I wonder whether it is possible to train a model on the 'JPG to camera model' problem.
"or makes up some bogus metadata."
Hm, I thought the Eiffel Tower clone in Japan was in Shanghai. The GPS coordinates seem to indicate that Tokio has one, too. Let's google to check that.
OK, that's more for the hacker engineer, but I thought one would not have to make that explicit on HN.
Umm, No, many popular photo-sharing sites at least give you the option to strip metadata from photos when posting. IIRC, fb does this as well, by default.
> and check whether there still are parts of the file with metadata in the disk's free blocks.
Good luck.
>A Nikon F5 at f/5.6 and 1/100s? No way!
It doesn't have to be believable metadata to be practically useless.
>Also, the aberration looks more like that of an Canon lens, but I'm not 100% sure of the model.
> I wonder whether it is possible to train a model on the 'JPG to camera model' problem.
Sure, but not for beer money. You could probably even have some success identifying individual cameras by characterizing their CCD/image sensor, and comparing to known photos.
There are no digital yearbooks from the high school. I checked
And then scrolling thru all the friends lists of all their friends on Facebook.
(Especially if they have nicely sorted their friends into "High School" friends.)
Twitter lets you do the same thing.
I've seen entire classes all "Friend" and "Follow" each other.
At least that's what I might have done.
Simplest explanation: He knew before he made the bet.
Of course one of the first things I did was ask another friend who I thought met her. But no, ultimately the answer did come from the internet, not from anyone I knew.
- This isn't the actual photo but looks similar, except that she was alone. Justin was afraid of reposting the same photo after the results of handing over a photo last time :)
- There was no useful exif data in the image.
- We knew her high school and approximate age in Austin, but were unable to find any yearbooks / class rosters online. (Idea not pursued: try to acquire a yearbook offline).
- Facebook Graph Search and LinkedIn with the criteria we knew didn't help.
- Google Image search didn't help.
> "I'm sure there's a lot of information missing from this story"
Quite true.
Ultimately we had several methods going at once to try to figure it out.
The one that came back with results first relies on some information Justin still doesn't know we had (though he would still be impressed), but other approaches we also expected to work would be possible with only what's in the story, and haven't yet been mentioned here.
EDIT: shared more at https://news.ycombinator.com/item?id=6369751
At that point, if the bartender knows some of her backstory or friends or anything - maybe only has a first name - things get much simpler.
The "extra" information we got came later after we already started working on it, and came as surprise. It just happened to bring the result back sooner than the other approach.
If you don't mind me asking, would other approaches discussed here have brought the result nonetheless? or did they?
If Justin observed correctly, you guys were very impressed with yourselves. Had you retrieved the answer by methods of which you were confident, you would've still been impressed, but not greatly.(I'm making a lapse conjecture here, of course)
Since you were confident that you would be able to find her, prior to looking at the picture, according to the article, I don't think you were looking for any information in the picture other than for confirmation.
However, some extra information in the picture(since it was taken very recently) enabled you to look further, which returned a result. Perhaps a building, an address, or some sort of landmark-ish location which narrowed whereabouts of the picture vastly? If this were the case, then I can understand that you guys were very impressed, because well, it is.
EDIT: Just saw the 'solution'; very nice touch!
So it seems you weren't even confident when you got your confirmation, a.k.a., it was an informed guess but could have just as easily been a girl that looked alike at an event at the time she was supposedly at an event?
Tricky ol' Phil.
* Everyone was at the original bar at the same time - someone else had photos from the night that had her and other people in them, and those people (her friends) could be found online.
* The bar had some kind of "sign-up for x" list, or some kind of contest, or some kind of photos of the night, which were helpful.
* You searched twitter/facebook/whatever for a girl posting about a boy she just met at bar X, blah blah
* We know she was going to some kind of event - you guys figured out what that was, and found a guest list or something.
* You guys did a some wi-fi sniffing and read a few emails. He did say he emailed you the photo, so you know which account to listen for.
Are any of those close?
OK then, there is something identifying about this girl, which the second photo gave you
* Tattoo
* Glasses
* Hair
* Amputee
* Scar/burn, etc.
* particular purse/hat/watch/accessory
* etc.
A credit card slip with her name on it perhaps?
I'm not sure what the rules are concerning tagging suggestions, but I have noticed that it suggests that I tag people I know in other people's photos so you KNOW they recognize everybody in every photo - the trick is meeting the right conditions for it to allow it to share that recognition match with you based on her privacy settings I think...
It is also possible that during the few days' correspondence one of the engineer guys had some other piece of info that we are not familiar with. Seeing how the engineers already knew her high school, I'm sure Justin talked about her with co-workers.
I think the key information that's missing here is what the engineers knew but Justin didn't know that they knew.(Sounds like a quote out of a Friends episode, mmm)
Or maybe you knew that she wanted him to meet her at an event and you found a roster for that event- or ie if it was a real estate convention you would have a pretty good idea that she was a realtor and could look at websites for realtors in Austin.
Or maybe you checked his recent facebook friends, looked at his recent twitter followers, and everyone who liked/commented on his recent photos/posts on Instagram and whatever other social networks he's on- as well as the photos/posts he liked/commented on.
I think you took a (relatively) low-tech approach to this.
- Create an ad on Facebook with her picture.
- Target people that went to her high school within 4 years of the range of years we thought she graduated within. Was only a few hundred people so the cost of running the ad would likely only be a few bucks.
- The ad copy / landing page just needed to be convincing that we weren't stalking her. We went with the "help us win a bet" approach but the "lost camera" approach would have been good also.
Ultimately we got the answer sooner from another approach after we found out that we had another picture of the girl that Justin didn't know we had. So we canceled the ad, but I think this would have worked if given enough time.
No, we weren't. And she knew about it, was amused by it herself. No bad intentions.
stalk. slang defination: hound, spy, solicitation for a certain kind of information
But you knew that.
her opinion and agreement apparently doesn't matter, the internet knows better.
Find the event Thursday night at 8:00pm in Austin and start tracking down from there.
Her dress, which was very unique had never been worn by her, because she bought it the day before
...should narrow down the events that this person wanted to attend. Here are the events that happened that day:
http://www.austinchronicle.com/calendar/2013-09-05/
...which, to me, only had one really unique event that evening:
http://www.facebook.com/events/1405498132998020
...which, possibly, would include a picture of that person either in the same outfit or include a profile picture of that person commenting or liking the event.
But this is just a guess...
The second photo you realized you had showed the subject holding the device used to take the picture sent to your colleague. You investigated that particular device and found a way to uniquely identify it's pictures based off some characteristic of it's photos, not necessarily EXIF, then scoured the Internet for photos with the same unique characteristic until you found one that revealed more information about the subject?
If your engineers as self-satisfied about finding her as it sounds, then they'll be happy to describe their process over a beer or two. I've never known an engineer who doesn't like discussing how they solved some riddle.
Option 1: Step 1) Find copies of yearbooks going back a few years, 2) find her face in the yearbook.
Option 2: Step 1) Search facebook for people who went to the same high school as she did in the probable time frame. 2) Look for her in that set, or in the set of friends that the people who matched the search have.
Feel free to sprinkle in face detection on the yearbook/facebook photos to make it more engineer-ey.
As a matter of fact, how did they find out what high school she went to? I can't imagine you telling them that if you were not providing much information about her to anyone else.
FYI, I am a sales guy...not an engineer
As someone who has been a software developer for quite a long time, this statement makes a big assumption.
A good software developer knows it is often far easier and faster to flip through a couple of hundred photos and process them with your meat-brain rather than write a one-off bit of software to do the matching, even if you start from an existing base like OpenCV or whatever.
That's why I said "Feel free to sprinkle in face detection on the yearbook/facebook photos to make it more engineer-ey."
If the photo was taken at her home and there is any landscape in the background, the obvious initial point of attack would be to identify the location the photograph was taken.
That's only a couple hundred kids, each with a couple hundred friends.
If you still get nothing, start doing friend requests for the people that don't show their friends. I'll bet you'd get another 25-30 kids that way.
I don't see why you'd need facial recognition software to go thru 20,000 pictures.
3/4ths of them would be the wrong gender/age and be immediately disqualified.
But if you're a programmer, you can probably scrape and aggregate to speed things up - drop out the males, other high schools, only show pictures once, etc.
The inside of her thumb is visible through the glass, so with enough detail it might be possible to record her thumb print. Even if you could do that, though, odds are against her print being present in any publicly-available database.
So, here's a thought. Let's say the Engineer has access to the SMS logs (NSA-style) for Justin's phone. (Maybe his phone goes through their system. But even if not, maybe it's company-provided, so there's an admin interface provided by the telecom that lists this information.)
Justin probably doesn't regularly text people in Texas, since the team is based out of Palo Alto. So look through his logs for a number in a Texas area code (http://en.wikipedia.org/wiki/List_of_Texas_area_codes) that he's been texting more often than usual.
That gets you the girl's phone number; probably a cell phone. Now, you need to go from the cell phone number to a last name. Reverse Phone Lookup probably won't work for this, although it's certainly worth a first try. More likely candidates: looking through the users table at close.io for a matching phone number, or doing a google search for the phone number. Or, try calling in the middle of the night and seeing what the voicemail says. Any of these approaches might work.
Once you've got a last name, grab a copy of the white pages. Most public schools publish their district boundaries, so go grab a copy of that. Look through all entries that match the last name, and see how many of them fall within (or close to, in the case politics made the boundary change) the district boundary. You probably will get a handful (maybe three or four). If you only get one, bingo!
If you get more than one, I'd call the cell phone to see if I can get a first name somehow. Then, I'd go through the landlines in the White Pages and call them one at a time, "Hello, may I speak to X? Sorry, wrong number." until one matches.
I did have an idea that we could try to hack his Verizon account to look at his SMS log, figuring that Forgot Password questions might be easy or that we could intercept an email. But we didn't do this.
I find it very likely that she was found on the internet somewhere.
1) They were running software on a computer for hours
They are engineers!
2) The first couple of tries didn’t work
Are you sure they weren't debugging?
3) They knew what high school she went too prior to searching
If they know the high school they only have to find one or two established teachers who were around for a long period of time. It is not unlikely they can call a name from a picture after all those years.
This smells like viral marketing to me (too bad I still don't know what close.io is, should have added some information to the article. Or are you looking for new engineers? "Hey look! At close.io we seal bets with engineers. About girls! Come work for us!").