18 Million Face Database Now Operating In The U.K.

Video and article about a 18 million strong FaceBase that was quietly brought online in the U.K. Some controversy since citizen's mugshots were used to populate the system, whether they were later convicted or not.

Ballsy quote:

[Ramsay] said searches of the database using facial recognition were 100% reliable in cases where there were clear images, and could be completed in seconds.

Maybe they will created a realtime method for matching faces from private security cameras. This would allow for personalized deterrence messages to be generated during an attempted break-in, which could be quite effective. Something like: "Hey Jimbo, you may want to think this over; with your two priors you are looking at 5 years, hard time". ;)


'100% reliable' is brazen spin. The '100%' is, almost by definition spin and 'reliable' seems purposefully vague, as 'accuracy' is clearly the issue.

It doesn't appear that the technology is new here, just the expansion of the application.

The reality is, across 18 million people, most are going to have a lot of other people that look surprisingly close to them.

The reality is, across 18 million people, most are going to have a lot of other people that look surprisingly close to them.

Unlike an access control application that would struggle with even a fraction of that size database, certain crime applications are more tolerant of multiple matches due to similarities. Consider when the local news media shows a sketch or picture of a suspect during a broadcast: They may get dozens of 'matches' called in by the public. But they do it because it's still a useful way to get leads, even if most are just mistaken identities.

In a similar way, this system is like being able to broadcast a picture for every crime, and if they limit the searches to ones where they have decent crime scene images, it seems quite useful.

Automated face recognition against the general population is the "flying cars" of video surveillance. Everybody thinks it's an inevitable future, but in spite of the fact that the technology does exist, it's simply not practical in application.

Face recognition would make a fine tool for (say) sifting through thousands of mug shots to match up an image of an unknown perpetrator to some candidates in what is really a large list of suspects (people with priors). But when using it to scan general population it's a poor fit.

When people see a police sketch of their local serial rapist on the news, they not only mentally match it up to others they know who the sketch reminds them of, but they also consider the character and situation of those people. So, a police sketch might make me think of my cousin's new boyfriend who I know is jobless and is using drugs. But it won't bring to mind my parish priest or my good buddy family man friend who I know was out of town on vacation during the time of the crime. Automated face recognition doesn't have the benefit of those other factors to help it narrow down it's search--rather it matches on anything that looks like the candidate image and is too easily taken down the wrong path by lighting conditions and differences in image quality that cause more problems than they're worth.

Face recognition would make a fine tool for (say) sifting through thousands of mug shots to match up an image of an unknown perpetrator to some candidates in what is really a large list of suspects (people with priors).

Right. That's what they are doing. They are not talking 'gen pop'.

Automated face recognition doesn't have the benefit of those other factors to help it narrow down it's search--rather it matches on anything that looks like the candidate image...

Just the opposite actually; since these are mug shots, they contain copious amounts data that can be used to limit the dataset for the actual facial recognition fuzzy image search. For instance, with your example of a local serial rapist the first order data set might only contain mugs of people who:

  1. Live within a 50 mile radius
  2. Weren't in custody at the time of the crime
  3. Have been arrested for rape in the past
  4. Meet the general Height, Weight, Skin Color, Age, Sex

What might that bring 18 million down to? 5000? 10000?

Or maybe search all mugs shots where ALPR had recorded a matching mug's plate at the approx. time and the place of the crime.

I think the main limitation is the quality of the photo from the crime, which obviously can vary greatly. Many pics are probably not worth the trouble today, but with the number of cameras increasing as well as their image quality, its value is growing everyday.