I was able to see a presentation about this topic one time by ICE. It tragic in two senses:
1) The crimes themselves are absolutely intolerable to society.
2) The lack of serious resourcing the seriousness of this crime should warrant.
Part of the presentation was specifically on identifying the location a sequence of photos was taken in the hopes that it would narrow down the search radius for the offender. The team that did the work were members of the ICE HERO program -- one of the coolest veteran-to-civilian programs I've ever seen [1][2]. The folks literally spent weeks "driving" around likely areas in Google maps trying to identify the location. Because the photo was taken from an angle that the car wouldn't have seen, they weren't able to use any automation at all and it literally took a human's intuitive understanding of 3-D spaces to eventually figure it out.
However, what was really tough was that the HERO program only exists because it offers an internship that costs very little to DHS -- they simply can't afford a large group of full-time staff.
I urge anybody who can to petition to better fund anti-child abuse programs.
How does the effectiveness of tracking down individual cases in this way compare to e.g. improving the eoconomic incentives overall? Commercial child abuse is mostly a result of deprived economic circumstances, in which people see more benefit in selling their children rather than giving them the best future possible, if I'm not mistaken.
I don't intend to spoil the enthusiasm here, but wouldn't supporting welfare hence be a more effective measure than investing in fancy tech? That latter only increases the deterrants which seem hard to increase any further to begin with.
Economic incentives aren't going to encourage someone to not sell their child for deplorable purposes; if you don't emotionally care for your child, money isn't going to make you care. On the other hand, if you can use a non-profit to pay terrible/uncaring/abusive parents to hand over their children voluntarily (surrendering parental rights) and to get sterilized to prevent the harm of any more children, that is a cause I would write checks to all day long.
Disclaimer: I am a parent, and have fostered neglected children.
One of the things I learned in the presentation is that economic incentives are not typically the motivation for this kind of evil. The "currency" in the environment appears to very much be fresh photos and videos of kids being sexually abused.
The trade in the material is unfortunately large. The descriptions of how the systems worked reminded me very much of the kinds of upload/download ratio systems you might find on software and movie/music piracy boards.
https://www.thorn.org/ (non-profit) is working hard to provide tooling to law enforcement to address the issues you mention, and posts on HN's Who's Hiring threads (front end engineers) [1].
I don't know if they accept volunteers for code contributions, but I hope they consider it if they already don't.
Is there not a volunteer program people can help with in this situation.
I'm a British citizen but looking at spending 2 or 3 months out in Cambodia working on coding projects an exploring a country I've never visited.
Unfortunately, I know Cambodia has had problems with child abuse and child sex tourism in the past. Working on something like you've just described seems like a fantastic way to explore a country and also track down nonces
How about instead of rounding up all sorts of harmless non-white people and throwing them in cages that instead their efforts are focused on the abhorrent things like that?
> If they did what risks and repercussions do they face?
Based on my experience as a victim of a crime, complaint to police very quickly reduced my risk of repercussions -- as soon as perpetrators faced police.
Your anecdotal experience is of only minor consequence here.
I've known people that have been abused by the police and other authority figures.
Plus if you're a minor and you're being abused by your guardian and the police show up to ask questions you're going to face backlash and it isn't going to be good.
For some there's really nowhere to go for justice.
Child sexual exploitation is a well researched bit of criminology. Jessica Eaton has a lot of research focusing on female victims which covers all of this stuff.
I've been trying to figure out what dennisgorelik's angle is in his comments in this thread and can't quite figure it out. I've decided not to respond directly because they're obviously not coming from a rational person who's spent even a few moments thinking about what he's writing.
Everything I've written as a response comes off as patronizing at best and incredibly rude at worst so I've decided to just not respond at all. It's kept me up half the night to be honest because his perniciously ignorant responses are so rage inducing.
I strongly agree with you, and it took me considerable effort to write the post I did as calmly as I did. And I'm going to take the advice to dis-engage.
> How do you even know from that story that there was sex or child involved? You are eager to engage in a manhunt based on ... what exactly?
Trust. We can see often cases of americans executed by traffic police just because, and other abuses ("A Dallas police officer entered an apartment she thought was hers and killed the owner of the apartment" and so...) so the repeated message that encounters with US police should be navigated carefully if you don't want to be another victim looks justified. At least in some places and specially if you are black or latino.
I think that in Europe we have a different relationship with the police. It seems (from my non expert point of view, I could be wrong) that Police in USA can lie and "get away with murder" anytime they want. Europol lying about such serious theme as abused children would be a big no. Would nuke any trust and public collaboration and is just not expected. What happened to Ian Murdock, Aaron Swartz or Andrew Sadek [1] would be very difficult to justify from an european point of view
> Without any details about what actually happened
I agree that is frustrating not having more details that could help us to link facts or notice subtle details that otherwise would be missed, but I can see how showing forever specific details about his/her sexual abuse in internet would do more harm than good to the victim.
> Without court decision?
You can't start a trial without having even the name of a suspect. It seems that is too early for that if they are compiling facts and trying to reach victims.
So somebody took a bunch of photos of children in public places. These children probably are not even aware about these photos and do not care either way. But activists from this post would like to turn these children into victims and drag them into dirty investigation. Then put the photographer into prison for 5-20 years and permanently label him "sex offender" for the rest of his life.
Do I understand the goals of this search correctly?
This is when I spare a thought for people whose job is to moderate user-posted content.
None of these pictures have signs of immediate physical violence, but just looking at them and thinking of their possible backstories makes my stomach churn.
Kind of disappointed in how few images there are there, considering the scale of child abuse mentioned in the article it is odd how there is only a dozen or so images to look at and try to identify.
The thing on the back looks surprisingly similar to a package of "Yerba Rosamonte"[1], which would place this picture in Argentina/Uruguay. But it's too low res to say whether that's the case, or just pareidolia.
I thought that container looked very familiar. My first guess was going to be Jergens, but I can't find an exact match on Google Images. Nivea might be it; again I can't find a dead-ringer but it's close to some of their packaging designs.
Unfortunately it's a global brand so not much help.
The context makes these cut outs really upsetting to see in some cases. Like the hyper zoom-ins of dolls and figurines in what may be a child rape scene.
Yup. I can see the pictograms for center 中 and strength/power 力. Maybe something related with healthcare or a gym.
The most interesting part are the palms, but they are damaged with the photoshop work. If they are what they seem, this particular kind of palms are picky and do not tolerate any frost. A city near the coast probably.
The blue trafic signal is also interesting.
(Could be this photo related with the coca-cola red seats?)
Remember when 4chan weaponized agains Shia LeBouf's flag and were able to determine the location of the ground flag based on the contrail evidence of the flight path of what the camera could see. The camera was only pointing at the sky, with the flag in the image. They were able to determine approx latitude due to the live stream revealing the sunset - and then comparing the paths of planes seen in the camera with flight data...
Then they drove to the location and took out the flag
I am not one to doubt the impressive capablity of a hoard of highly motivated 4chan autists, but it does become a lot less spectacular of a feat if you consider LaBeouf was spotted in a Greenville, TN diner around the same time as the flag hunt was underway (which was twitter knowledge). I thought their triangulation efforts through driving around, beeping a car horn, and listening if the sound came through on cam audio was the most impressive part of their detective work.
But that was a live video feed in a known country where they could cross reference weather conditions with commercial flight information (and rumors about LeBouf sightings). This is just a static picture that was taken somewhere on planet Earth.
Using "only" North Korean propaganda videos and commercially available satellite images and published memoirs from NK defectors, they were able to located the building where NK is making their TELs.
The people following the Syrian civil war are rather dedicated and the training camp was not the first or last time some videos got geolocated (though without any real life consequences in the other cases). I dont see a reason for a PR stunt as you describe it.
I'd bee willing to join an IN team trying to do this ... If you can string together a few non-traditional clues to the general area you could target the crowd-sourced part more effectively. Might even be able to get cloud providers to donate unused CPU.
That's an interesting design of swingset; not one I've seen before.
Unlike the bog-standard swingset in the US, which is typically all steel tube, that one looks like it has extruded (aluminum?) verticals and then a tubular horizontal weldment as the top piece. Further, the horizontal looks like it was hot-dip galvanized as a finished unit. The verticals were painted before assembly, not after, because of the bare-metal bolt heads. The fasteners are presumably stainless, since I don't see any rust marks below them. It's got an interesting bearing/bushing at the top of each chain, probably to make it not squeak; that strikes me as a fairly luxe feature.
I couldn't find any similar ones searching around online, although my Google searches don't turn up many European models at all (thanks, Google).
The photo doesn't go up enough to be sure, but it looks like the top bar is an unbroken one-piece section. That would make it pretty bulky and awkward to ship. (The vertical pieces look like they come apart into about 4 ft sections, which means they were likely palletized at one point. Maybe the extrusions are not done in the same place as the rest of the manufacturing..? That's certainly plausible.)
If someone recognized the design/model, I think that would probably be a significant clue towards the photo's location; at the very least, the manufacturer would probably have a fairly good idea of where they normally sell to.
It has the swept-bend tubular corner reinforcements, the polyethylene (or some other kind of plastic) panels on the upper corners, etc. The only difference I can see between the product photo and the Europol one is that, in the Europol photo, the top bar seems galvanized rather than painted. But that's the sort of thing that could easily change as a product is marketed.
It appears the company is out of Krasnodar (Краснода́р), located in southern Russia.
The metal arches as reinforcements in the top corners are not so common. The use of polyethylene plaques as decor and reinforcements started in 1970 it seems.
I wonder if people who use a popular photo sharing site like Google Photos would be in favor of there photos with location meta data being used for a project like this.
I think I would if I could trust that it was only being used for something like this & not commercial advertising type projects.
Sure. But I think Pizzagate has shown us that there's enough weirdos out there who'll show up to a place demanding answers.
And don't forget, people are strongly died to locations. If Reddit decides that a particular photo is located at 287 FM 772, that's someone's house, and a name can easily be associated with it. Nobody wants to receive a Google alert that their name is suddenly popping up in /r/find-a-pedophile
What incentive could be added to have people view these photos on a regular basis? I think facebook could be one answer with it's global reach. They could do an official page or something and offer real discounts at partnering restaurants or stores from around the world if you spend time each month viewing all the pictures. Imagine the reach they could have. At the same time I feel that there should not be a business of restaurants and the like selling coupons on abuse photos but I think the greater good of finding these kids may be there. The reason I add a type of payment is because it is my belief that though people do mean well, without an incentive they are less likely to view these images. Or perhaps if add companies put those kind of pictures on the websites instead of ads I would disable the adblocker.
Photo number 9. The box shows a picture of an Asteraceae that fits well with Cichorium intybus, a vegetable cultured in temperate areas (from Europe mainly), and the colours of the Romanian flag.
Can't identify the background trees at this low resolution, but there is a small Fabaceae tree in the photo that looks like either Robinia pseudoacacia or Sophora. This trees have heavy seeds that do not go very far. Some bigger trees from one of this species could probably be seen near this area. Maybe in the streets (or in a garden).
Just realized that is the same area as this second photo, so if is Asia nothing of the former is appliable. The list of possible fabaceae trees is much bigger then. There is a pinnate-leaved palm also, so in principle areas colder than -12 could be excluded (USDA 8a).
Ok, I was looking at the medium size. At larger resolution the photo is more interesting that it seems. Four different threes and something more. They should examine it more carefully.
The big trunk in the background is from the same Fabaceae also. Senna could show a similar bark and leaves. Senna siamea from Thailand for example looks similar. There is a sapling of something that looks like a Psidium also and maybe a Schefflera. All tentatively.
In this photo there is a small tree at rigth that is partially clipped. Compound leaves with five leaflets. Probably Fraxinus. Fits well. Temperate area in this case.
It would be great if that was possible. But in the meantime human beings are amazingly good at this kind of thing. Our brains are highly optimised for this kind of work.
Google earth was used to narrow down one of the locations (says so on the video) but required some serious sluething from an architect, so I presume beyond the means of current AI tech.
Hopefully we'll come to a point soon where AI and deep learning can help out in this space. I remember models being trained that could predict the country based on a street view photo with a high degree of accuracy, which would be a good first step. The next one would be to either match a top-down satellite photo with a picture, or figure out a way to match street view photos based on similarity.
Given that Google actually has the data as well as the skills and computing power to help out Europol, I'd have hoped they'd be jumping at the opportunity to do so. At the least it would make for great PR, but perhaps they're afraid of the possible privacy issues and discussions as well (I can already see the headlines: "Google knows the location of any yard appearing in a photo!").
In any case, it's telling how humans are so good at this stuff and picking up on clues and patterns. It'll still take a while before machines are as good.
That’s not really how any of this works—you can’t simply throw a bunch of unrelated data at some algorithms and expect usable output. (There’s also no such thing as “deep learning”.)
Law enforcement and related organizations already use machine learning quite a bit, particularly for image enhancement. Yes, Google does have a lot of images of various locations from a top-down perspective, but that isn’t helpful for accurately determining a location from the images that Europol collects.
Also, keep in mind that Europol is only posting images here when all other means of determining location and identity have been exhausted. The images are usually indoors and don’t contain enough information for ML to be of any significant use. You might be able to narrow it down to a probably country based colors and design patterns, but that’s hardly sufficient and not solid enough evidence to actually do anything.
> That’s not really how any of this works—you can’t simply throw a bunch of unrelated data at some algorithms and expect usable output.
That is never what I claimed. First, note that I took a pretty (IMO) balanced view and indicated that this is still a hard setting. Second, note that I did indicate that sufficient training (i.e. labeled) data would be required.
> Yes, Google does have a lot of images of various locations from a top-down perspective, but that isn’t helpful for accurately determining a location from the images that Europol collects. You might be able to narrow it down to a probably country based colors and design patterns, but that’s hardly sufficient and not solid enough evidence to actually do anything.
Maybe not completely, but again: being able to narrow it down would already be an incredible help, especially for outdoor pictures (which were also shown in the article's video). I never claimed that a model would completely replace the human process.
Also, I find the downvotes (not saying you) on my initial comment to be in pretty bad form. I'm not Jeremy Howard or Andrew Ng, but don't think I was blowing smoke, and work in the area of data science and ML.
I was able to see a presentation about this topic one time by ICE. It tragic in two senses:
1) The crimes themselves are absolutely intolerable to society.
2) The lack of serious resourcing the seriousness of this crime should warrant.
Part of the presentation was specifically on identifying the location a sequence of photos was taken in the hopes that it would narrow down the search radius for the offender. The team that did the work were members of the ICE HERO program -- one of the coolest veteran-to-civilian programs I've ever seen [1][2]. The folks literally spent weeks "driving" around likely areas in Google maps trying to identify the location. Because the photo was taken from an angle that the car wouldn't have seen, they weren't able to use any automation at all and it literally took a human's intuitive understanding of 3-D spaces to eventually figure it out.
However, what was really tough was that the HERO program only exists because it offers an internship that costs very little to DHS -- they simply can't afford a large group of full-time staff.
I urge anybody who can to petition to better fund anti-child abuse programs.
1 - https://www.ice.gov/hero
2 - http://www.herocorps.net/
How does the effectiveness of tracking down individual cases in this way compare to e.g. improving the eoconomic incentives overall? Commercial child abuse is mostly a result of deprived economic circumstances, in which people see more benefit in selling their children rather than giving them the best future possible, if I'm not mistaken.
I don't intend to spoil the enthusiasm here, but wouldn't supporting welfare hence be a more effective measure than investing in fancy tech? That latter only increases the deterrants which seem hard to increase any further to begin with.
Economic incentives aren't going to encourage someone to not sell their child for deplorable purposes; if you don't emotionally care for your child, money isn't going to make you care. On the other hand, if you can use a non-profit to pay terrible/uncaring/abusive parents to hand over their children voluntarily (surrendering parental rights) and to get sterilized to prevent the harm of any more children, that is a cause I would write checks to all day long.
Disclaimer: I am a parent, and have fostered neglected children.
projectprevention.org pays serial child neglecters to get birth control
> Disclaimer: I am a parent, and have fostered neglected children.
And that makes you an expert in criminology?
I don’t proclaim to be an expert, just someone with an opinion and some domain experience.
You need to rely on large scale studies to make statements about the population at large.
You can use your small number of data points (anecdotes) to generate hypothses, but you need a lot more data to test these hypotheses.
Build the study and I’ll find someone to fund it. No one ever said, “we have too much good data”.
One of the things I learned in the presentation is that economic incentives are not typically the motivation for this kind of evil. The "currency" in the environment appears to very much be fresh photos and videos of kids being sexually abused.
The trade in the material is unfortunately large. The descriptions of how the systems worked reminded me very much of the kinds of upload/download ratio systems you might find on software and movie/music piracy boards.
https://www.thorn.org/ (non-profit) is working hard to provide tooling to law enforcement to address the issues you mention, and posts on HN's Who's Hiring threads (front end engineers) [1].
I don't know if they accept volunteers for code contributions, but I hope they consider it if they already don't.
[1] https://news.ycombinator.com/item?id=19281914
Is there not a volunteer program people can help with in this situation.
I'm a British citizen but looking at spending 2 or 3 months out in Cambodia working on coding projects an exploring a country I've never visited.
Unfortunately, I know Cambodia has had problems with child abuse and child sex tourism in the past. Working on something like you've just described seems like a fantastic way to explore a country and also track down nonces
You can take photos of every hotel room you visit. These get uploaded to a database usable by law enforcement: https://traffickcam.com/
Europol has a "trace this item" website: https://www.europol.europa.eu/stopchildabuse
Traffickcam is a fantastic idea I'll certainly use that.
Just re-read that.
I don't mean to view actual child abuse images but similar to the europol thing ie the actual victims deleted from the images.
How about instead of rounding up all sorts of harmless non-white people and throwing them in cages that instead their efforts are focused on the abhorrent things like that?
These "presentations" are very vague. What crimes, specifically, do they try to prevent or prosecute?
Why victims are not complaining about these crimes?
If they did who would listen? If they did what risks and repercussions do they face?
This also presumes they're strong enough to speak up in the first place. If you're a victim of this kind of abuse you usually aren't.
> If they did who would listen?
Police.
> If they did what risks and repercussions do they face?
Based on my experience as a victim of a crime, complaint to police very quickly reduced my risk of repercussions -- as soon as perpetrators faced police.
> this kind of abuse
What kind of abuse?
Police. Right.
Sure.
Your anecdotal experience is of only minor consequence here.
I've known people that have been abused by the police and other authority figures.
Plus if you're a minor and you're being abused by your guardian and the police show up to ask questions you're going to face backlash and it isn't going to be good.
For some there's really nowhere to go for justice.
> I've known people that have been abused by the police
I've been abused by the police.
That did not prevent me from complaining to police when I got in trouble with criminals.
> For some there's really nowhere to go for justice.
Correct. But many victims still have ability to get justice. And it is very hard for potential perpetrator to predict if victim would complain or not.
So that uncertainty works as an effective deterrent, even if some of the victims are not going to complain.
If you're white and male, sure, you can go to the cops and they may or may not do something depending on where you live.
If you're not they may think it's just hilarious to torment you, to ridicule you, or worse.
In some places in America, for many people, the police are the enemy and cannot be engaged with in any capacity for any reason whatsoever.
> What kind of abuse?
Child sexual exploitation is a well researched bit of criminology. Jessica Eaton has a lot of research focusing on female victims which covers all of this stuff.
I've been trying to figure out what dennisgorelik's angle is in his comments in this thread and can't quite figure it out. I've decided not to respond directly because they're obviously not coming from a rational person who's spent even a few moments thinking about what he's writing.
Everything I've written as a response comes off as patronizing at best and incredibly rude at worst so I've decided to just not respond at all. It's kept me up half the night to be honest because his perniciously ignorant responses are so rage inducing.
I strongly agree with you, and it took me considerable effort to write the post I did as calmly as I did. And I'm going to take the advice to dis-engage.
> can't quite figure it out. I've decided not to respond
"No communication" attitude does not really help with figuring out what other people think.
> obviously
"can't quite figure it out", but conclusions are "obvious" anyway?
> It's kept me up half ... so rage inducing
That hints that there is an unresolved conflict.
How do you know if there was exploitation if there were no complaints?
How do you even know from that story that there was sex or child involved?
You are eager to engage in a manhunt based on ... what exactly?
> How do you even know from that story that there was sex or child involved? You are eager to engage in a manhunt based on ... what exactly?
Trust. We can see often cases of americans executed by traffic police just because, and other abuses ("A Dallas police officer entered an apartment she thought was hers and killed the owner of the apartment" and so...) so the repeated message that encounters with US police should be navigated carefully if you don't want to be another victim looks justified. At least in some places and specially if you are black or latino.
I think that in Europe we have a different relationship with the police. It seems (from my non expert point of view, I could be wrong) that Police in USA can lie and "get away with murder" anytime they want. Europol lying about such serious theme as abused children would be a big no. Would nuke any trust and public collaboration and is just not expected. What happened to Ian Murdock, Aaron Swartz or Andrew Sadek [1] would be very difficult to justify from an european point of view
[1] https://www.thedailybeast.com/student-drug-informant-found-w...
"Trust" without any details about what actually happened?
Without any details about where these photos came from?
Without court decision?
> Without any details about what actually happened
I agree that is frustrating not having more details that could help us to link facts or notice subtle details that otherwise would be missed, but I can see how showing forever specific details about his/her sexual abuse in internet would do more harm than good to the victim.
> Without court decision?
You can't start a trial without having even the name of a suspect. It seems that is too early for that if they are compiling facts and trying to reach victims.
> where these photos came from?
[shrug] pedophilic archives in internet?
> pedophilic archives in internet?
So somebody took a bunch of photos of children in public places. These children probably are not even aware about these photos and do not care either way. But activists from this post would like to turn these children into victims and drag them into dirty investigation. Then put the photographer into prison for 5-20 years and permanently label him "sex offender" for the rest of his life.
Do I understand the goals of this search correctly?
well because the victims are usually threatened, children.
"Usually" does not mean "always".
Giving threats is a crime in itself.
Threatening victims does not guarantee victims' silence.
If that crime is widespread then there still should be plenty of victims who, actually, report that crime.
Here the direct link to the picture database https://www.europol.europa.eu/stopchildabuse
This is when I spare a thought for people whose job is to moderate user-posted content.
None of these pictures have signs of immediate physical violence, but just looking at them and thinking of their possible backstories makes my stomach churn.
Don't let your imagination run wild. Most child abuse is nonviolent, usually committed by family or acquaintances of the child.
Kind of disappointed in how few images there are there, considering the scale of child abuse mentioned in the article it is odd how there is only a dozen or so images to look at and try to identify.
I think they are releasing a few at a time to try and get the most eyeballs concentrated on the images.
I don't know if that's the most effective strategy, but maybe there are other reasons why they can't just dump them all at once.
https://www.europol.europa.eu/sites/default/files/styles/eur...
That card looks like a pokemon card with a Google Chrome logo. Frustratingly familiar.
The thing on the back looks surprisingly similar to a package of "Yerba Rosamonte"[1], which would place this picture in Argentina/Uruguay. But it's too low res to say whether that's the case, or just pareidolia.
[1] https://www.google.com/search?tbm=isch&q=yerba%20rosamonte&t...
It looks like something else once looking at the same picture with the adjusted light (using Photoshop):
https://ibb.co/1K1kZR1
I thought exactly the same.
That blue can looks like a Nivea 250ml can.
I thought that container looked very familiar. My first guess was going to be Jergens, but I can't find an exact match on Google Images. Nivea might be it; again I can't find a dead-ringer but it's close to some of their packaging designs.
Unfortunately it's a global brand so not much help.
It looks a bit like the Domino's Pizza logo (but the whitespace inside the blue area wouldn't bleed so much into the outside, I think) https://en.wikipedia.org/wiki/File:Domino%27s_pizza_logo.svg
Voltorb, maybe? https://www.pokemon.com/uk/pokemon-tcg/pokemon-cards/xy-seri...
The context makes these cut outs really upsetting to see in some cases. Like the hyper zoom-ins of dolls and figurines in what may be a child rape scene.
Image 18 should be simple enough if anyone here can read Chinese - the businesses across the road seem nearly clear enough to read.
Yup. I can see the pictograms for center 中 and strength/power 力. Maybe something related with healthcare or a gym.
The most interesting part are the palms, but they are damaged with the photoshop work. If they are what they seem, this particular kind of palms are picky and do not tolerate any frost. A city near the coast probably.
The blue trafic signal is also interesting.
(Could be this photo related with the coca-cola red seats?)
It would be cool if someone like Google could implement this in their recaptchas somehow to get more exposure.
Some of the techniques a group on 4chan used are interesting: https://www.ign.com/boards/threads/remember-when-4chan-users...
Remember when 4chan weaponized agains Shia LeBouf's flag and were able to determine the location of the ground flag based on the contrail evidence of the flight path of what the camera could see. The camera was only pointing at the sky, with the flag in the image. They were able to determine approx latitude due to the live stream revealing the sunset - and then comparing the paths of planes seen in the camera with flight data...
Then they drove to the location and took out the flag
Joe Rogan on it:
https://www.youtube.com/watch?v=AoBK1_ixiro
Some random on it:
https://www.youtube.com/watch?v=n-ABXvbMSc4
I am not one to doubt the impressive capablity of a hoard of highly motivated 4chan autists, but it does become a lot less spectacular of a feat if you consider LaBeouf was spotted in a Greenville, TN diner around the same time as the flag hunt was underway (which was twitter knowledge). I thought their triangulation efforts through driving around, beeping a car horn, and listening if the sound came through on cam audio was the most impressive part of their detective work.
But that was a live video feed in a known country where they could cross reference weather conditions with commercial flight information (and rumors about LeBouf sightings). This is just a static picture that was taken somewhere on planet Earth.
Geo-location is a fascinating topic.
Here's a Twitter thread where BBC solved a murder in Cameroon: https://twitter.com/bbcafrica/status/1044186344153583616?lan...
From a single video of the crime, they were able to figure out where, who, and when (!) the crime was committed. (Using Google maps.)
And here's Jeffrey Lewis presenting how they figured out "Where does North Korea Build Its Missile Launchers?" (1h video): https://www.youtube.com/watch?v=E9QbI8Jf4hg
Using "only" North Korean propaganda videos and commercially available satellite images and published memoirs from NK defectors, they were able to located the building where NK is making their TELs.
Does anyone have any other examples?
The Jeffrey Lewis talk was incredible, thank you for sharing.
Bellingcat does this and more - they also teach a course on it
they triangulated the location of the ISIS execution videos using nothing more than shadows and geographical mounds in the desert
https://www.bellingcat.com/tag/geolocation/
I wonder if the posts suggesting the actual coordinates and messages were made by people connected to russian intelligence.. Just a thought
The people following the Syrian civil war are rather dedicated and the training camp was not the first or last time some videos got geolocated (though without any real life consequences in the other cases). I dont see a reason for a PR stunt as you describe it.
I'd bee willing to join an IN team trying to do this ... If you can string together a few non-traditional clues to the general area you could target the crowd-sourced part more effectively. Might even be able to get cloud providers to donate unused CPU.
Looking at the swingset photo (https://www.europol.europa.eu/sites/default/files/styles/eur...)
That's an interesting design of swingset; not one I've seen before.
Unlike the bog-standard swingset in the US, which is typically all steel tube, that one looks like it has extruded (aluminum?) verticals and then a tubular horizontal weldment as the top piece. Further, the horizontal looks like it was hot-dip galvanized as a finished unit. The verticals were painted before assembly, not after, because of the bare-metal bolt heads. The fasteners are presumably stainless, since I don't see any rust marks below them. It's got an interesting bearing/bushing at the top of each chain, probably to make it not squeak; that strikes me as a fairly luxe feature.
I couldn't find any similar ones searching around online, although my Google searches don't turn up many European models at all (thanks, Google).
The photo doesn't go up enough to be sure, but it looks like the top bar is an unbroken one-piece section. That would make it pretty bulky and awkward to ship. (The vertical pieces look like they come apart into about 4 ft sections, which means they were likely palletized at one point. Maybe the extrusions are not done in the same place as the rest of the manufacturing..? That's certainly plausible.)
If someone recognized the design/model, I think that would probably be a significant clue towards the photo's location; at the very least, the manufacturer would probably have a fairly good idea of where they normally sell to.
Replying to myself because I got an email from a user with a match:
http://igra23.ru/katalog/karuseli-kacheli-balansiry/kacheli-...
It has the swept-bend tubular corner reinforcements, the polyethylene (or some other kind of plastic) panels on the upper corners, etc. The only difference I can see between the product photo and the Europol one is that, in the Europol photo, the top bar seems galvanized rather than painted. But that's the sort of thing that could easily change as a product is marketed.
It appears the company is out of Krasnodar (Краснода́р), located in southern Russia.
This is an exact fit. Great job
Looks like a Lappset to me. That does not help much, they are selling in at least 50 countries https://www.lappset.com/Contact-us
This is not it, but is pretty close (just found by google images on: "playground swings aluminum square posts").
[0] https://www.eibe.net/en/swing-kondor-51007201100.html?c=5010...
Is that an unusual swingset design? I'm pretty sure i've seen it in lots of places...
The metal arches as reinforcements in the top corners are not so common. The use of polyethylene plaques as decor and reinforcements started in 1970 it seems.
I wonder if people who use a popular photo sharing site like Google Photos would be in favor of there photos with location meta data being used for a project like this.
I think I would if I could trust that it was only being used for something like this & not commercial advertising type projects.
This should be a subreddit. Reddit is pretty good for these things, they have a big, global audience and a great comment system for such things.
e.g. https://www.reddit.com/r/whatisthisthing/ works great.
They also have a habit of mis-identifying people and hounding them down.
well this material seems luckily carefully selected and without I persons visible. A found place is also pretty easy to verify.
Sure. But I think Pizzagate has shown us that there's enough weirdos out there who'll show up to a place demanding answers.
And don't forget, people are strongly died to locations. If Reddit decides that a particular photo is located at 287 FM 772, that's someone's house, and a name can easily be associated with it. Nobody wants to receive a Google alert that their name is suddenly popping up in /r/find-a-pedophile
What incentive could be added to have people view these photos on a regular basis? I think facebook could be one answer with it's global reach. They could do an official page or something and offer real discounts at partnering restaurants or stores from around the world if you spend time each month viewing all the pictures. Imagine the reach they could have. At the same time I feel that there should not be a business of restaurants and the like selling coupons on abuse photos but I think the greater good of finding these kids may be there. The reason I add a type of payment is because it is my belief that though people do mean well, without an incentive they are less likely to view these images. Or perhaps if add companies put those kind of pictures on the websites instead of ads I would disable the adblocker.
What's wrong with money
Photo number 9. The box shows a picture of an Asteraceae that fits well with Cichorium intybus, a vegetable cultured in temperate areas (from Europe mainly), and the colours of the Romanian flag.
https://www.europol.europa.eu/sites/default/files/styles/eur...
over the green square, appears a whitish blur that could easily say Cicoarea in a stylised way (or maybe is just another picture)
This is also a good article, not sure if it's been featured here before. https://www.bellingcat.com/resources/case-studies/2018/11/08...
Can't identify the background trees at this low resolution, but there is a small Fabaceae tree in the photo that looks like either Robinia pseudoacacia or Sophora. This trees have heavy seeds that do not go very far. Some bigger trees from one of this species could probably be seen near this area. Maybe in the streets (or in a garden).
https://www.europol.europa.eu/sites/default/files/styles/eur...
Robinia is a fully hardy and widespread species that can be found in almost all Europe and USA, USDA area 1a.
Just realized that is the same area as this second photo, so if is Asia nothing of the former is appliable. The list of possible fabaceae trees is much bigger then. There is a pinnate-leaved palm also, so in principle areas colder than -12 could be excluded (USDA 8a).
https://www.europol.europa.eu/sites/default/files/styles/eur...
Ok, I was looking at the medium size. At larger resolution the photo is more interesting that it seems. Four different threes and something more. They should examine it more carefully.
The big trunk in the background is from the same Fabaceae also. Senna could show a similar bark and leaves. Senna siamea from Thailand for example looks similar. There is a sapling of something that looks like a Psidium also and maybe a Schefflera. All tentatively.
In this photo there is a small tree at rigth that is partially clipped. Compound leaves with five leaflets. Probably Fraxinus. Fits well. Temperate area in this case.
https://www.europol.europa.eu/sites/default/files/styles/eur...
If is Fraxinus excelsior (probably) Europe. If Fraxinus pensylvannica: Eastern half of USA or Argentina.
The writing on the backpack looks Cyrillic to me.
Andrew Sullivan used to run a contest doing something similar albeit less serious and with remarkably good results
http://dish.andrewsullivan.com/vfyw-contest/
If there can be a first sort done by AI and then human contribution to try to narrow it better it might be helpfull.
Like « this picture seems to be located in this area/country » and then people from the area could have a look.
I’m pretty sure people would like to help.
It would be great if that was possible. But in the meantime human beings are amazingly good at this kind of thing. Our brains are highly optimised for this kind of work.
Please no thanks!
Once AI is brought into equation, project executives put all efforts into removing Humans from the chain because AI is cheaper
See YouTube AI based moderation and infractions, it's too difficult and no where to go if AI slaps you
Do not reverse search images on Google.
Should be solveable with a NN trained on Matching Google Streetview and the images..
Google Street View is limited to what you can see from the street. It would not work for several types of location.
Very true, but the article says there are 40 million images. Likely will work for a non-zero portion of the images.
Google earth was used to narrow down one of the locations (says so on the video) but required some serious sluething from an architect, so I presume beyond the means of current AI tech.
Yeah and we shouldn't worry about climate change because AI will just magically solve it like Sergey says.
You recognize how naive stuff like this sounds right?
Hopefully we'll come to a point soon where AI and deep learning can help out in this space. I remember models being trained that could predict the country based on a street view photo with a high degree of accuracy, which would be a good first step. The next one would be to either match a top-down satellite photo with a picture, or figure out a way to match street view photos based on similarity.
Given that Google actually has the data as well as the skills and computing power to help out Europol, I'd have hoped they'd be jumping at the opportunity to do so. At the least it would make for great PR, but perhaps they're afraid of the possible privacy issues and discussions as well (I can already see the headlines: "Google knows the location of any yard appearing in a photo!").
In any case, it's telling how humans are so good at this stuff and picking up on clues and patterns. It'll still take a while before machines are as good.
That’s not really how any of this works—you can’t simply throw a bunch of unrelated data at some algorithms and expect usable output. (There’s also no such thing as “deep learning”.)
Law enforcement and related organizations already use machine learning quite a bit, particularly for image enhancement. Yes, Google does have a lot of images of various locations from a top-down perspective, but that isn’t helpful for accurately determining a location from the images that Europol collects.
Also, keep in mind that Europol is only posting images here when all other means of determining location and identity have been exhausted. The images are usually indoors and don’t contain enough information for ML to be of any significant use. You might be able to narrow it down to a probably country based colors and design patterns, but that’s hardly sufficient and not solid enough evidence to actually do anything.
> There’s also no such thing as “deep learning”.
What do you mean by that?
https://en.wikipedia.org/wiki/Deep_learning
> That’s not really how any of this works—you can’t simply throw a bunch of unrelated data at some algorithms and expect usable output.
That is never what I claimed. First, note that I took a pretty (IMO) balanced view and indicated that this is still a hard setting. Second, note that I did indicate that sufficient training (i.e. labeled) data would be required.
This is what was possible in 2016: https://www.theverge.com/2016/2/25/11112594/google-new-deep-...: "The new deep-learning program churns through millions of photos to determine the best match."
Also see project of a fast.ai participant: "Which of the 110 countries a satellite image belongs to?" (point 13 here: https://forums.fast.ai/t/deep-learning-lesson-2-notes/28772)
> (There’s also no such thing as “deep learning”.)
- https://www.deeplearningbook.org/
- https://www.coursera.org/courses?query=deep%20learning
- https://eu.udacity.com/course/intro-to-tensorflow-for-deep-l...
- https://www.edx.org/professional-certificate/ibm-deep-learni...
- https://www.deeplearning.ai
- https://www.fast.ai/
> Yes, Google does have a lot of images of various locations from a top-down perspective, but that isn’t helpful for accurately determining a location from the images that Europol collects. You might be able to narrow it down to a probably country based colors and design patterns, but that’s hardly sufficient and not solid enough evidence to actually do anything.
Maybe not completely, but again: being able to narrow it down would already be an incredible help, especially for outdoor pictures (which were also shown in the article's video). I never claimed that a model would completely replace the human process.
Also, I find the downvotes (not saying you) on my initial comment to be in pretty bad form. I'm not Jeremy Howard or Andrew Ng, but don't think I was blowing smoke, and work in the area of data science and ML.
Maybew Europol could do a Kaggle competition?