A promising data point for the durability of facial recognition biometrics

Facial recognition helped the US Marshals Service to identify a military fugitive who escaped federal prison in 1977 (Fox News)

As far as biometric modalities go, the durability of facial recognition has been the easiest to test. Scientists have had access photos of known individuals taken over long periods of time stretching back to the days of Abraham Lincoln.

Still, this real world example illustrates that facial recognition biometrics are useful even over very long time periods even without frequent database updates.

“On schedule, within budget and within scope.”

Next Generation Identification: A closer look at the FBI’s billion-dollar biometric program (Biometric Update)

Representing a $1.2 billion investment by the U.S. federal government, the FBI’s massive Next Generation Identification (NGI) program is a ten-year lifecycle project that hinges on biometric identification technologies and has seen privacy advocates butt heads with law enforcement since its inception.

Split into six “increments,” Lockheed Martin was awarded a contract in 2008 to design, build and implement the program on behalf of the FBI, which ultimately aims to enhance the abilities of the agency’s aging IAFIS from the mid-nineties.

More at the link.

Japan, U.S. law enforcement to share fingerprint databases online (Japan Today)

Japan and the United States have agreed to provide mutual access to online fingerprint databases to aid criminal investigations.

According to the arrangement, each nation will have instant access to fingerprint data for the purpose of investigating individuals suspected of involvement in terrorism or other serious crimes such as murder, Japanese officials said.

Face-in-a-crowd systems have improved, but aren’t quite ready

Facial scanning system for crowds shows improvement in tests by Homeland Security (ABA Journal)

The scanning system is known as Biometric Optical Surveillance System, or BOSS. It uses a computer to build a “3-D signature” from photos taken at different angles, then compares the results with a watch-list database. A computer professor and an official with the company developing BOSS said advances in computer processing will eventually result in speedier and more reliable matches. Currently, the system’s accuracy is only 60 percent to 70 percent at longer distances. And it takes 30 seconds to process the images.

Expect to see a whole lot more of this…

Chicago police go high-tech with facial recognition software (Chicago Sun-Times)

Story Image
Pierre Martin

Police had a photo captured on a CTA surveillance camera on Jan. 28 of a suspected mugger, looking to the side, after he had just allegedly stolen a cellphone from a man at gunpoint on a Pink Line train.

Police also had an ocean of photos for comparison — 4.5 million criminal booking shots.

They ran the program.

And Pierre Martin ranked No. 1 on a list no one wants to top.

We at SecurLinx deploy help police deploy these capabilities. The article is correct to note that officer training is a hugely important factor in successfully adding this capability to law enforcement efforts.

India to require fingerprints before issuing SIM card

Soon You Will Require Fingerprints To Buy A SIM Card (SiliconIndia)

To put an end to the unauthorized distribution and access of SIM cards by fraudsters, the home ministry has asked Department of Telecommunication to explore various measures to take biometric details including fingerprints by cell phone service providers before activating the connection.

India isn’t alone. Pakistan is considering a similar requirement for purchasing a SIM (Subscriber Identity Module) card as a way of more definitively tying mobile phones to their purchasers.

Nigeria implemented a similar system beginning in 2010.

There are several reasons that countries want to do this. Most are related to making it easier (or even possible) to investigate crime. Mobile phones are critically important tools in criminal enterprises such as ransoming kidnapping victims and organized robbery. Terrorists depend upon mobile phones both for communication and to detonate explosive devices: Tele-operators briefed on biometric system (The Nation)

“NADRA being the sole custodian of biometrics of over 96 percent total population of the country, has offered the biometric solution in the wake of Interior Ministry’s grave security concerns over the use of cellular devices in terrorist plots,” the spokesperson said. It should be noted that on December 1, 2012, the Prime Minister, after taking notice of insecure sales mechanism for issuance of SIMs, directed all telecom companies to employ biometric verification for SIMs issuance within two months’ time.

EFF sues for FBI response to FOIA request

EFF Sues FBI For Access to Facial-Recognition Records (Electronic Frontier Foundation)

As the FBI is rushing to build a “bigger, faster and better” biometrics database, it’s also dragging its feet in releasing information related to the program’s impact on the American public. In response, the Electronic Frontier Foundation (EFF) today filed a lawsuit to compel the FBI to produce records to satisfy three outstanding Freedom of Information Act requests that EFF submitted one year ago to shine light on the program and its face-recognition components.

Since early 2011, EFF has been closely following the FBI’s work to build out its Next Generation Identification (NGI) biometrics database, which would replace and expand upon the Integrated Automated Fingerprint Identification System (IAFIS). The new program will include multiple biometric identifiers, such as iris scans, palm prints, face-recognition-ready photos, and voice data, and that information will be shared with other agencies at the local, state, federal and international levels. The face recognition component is set to launch in 2014.

The text of the actual suit is also available at the EFF site [pdf] here.

Adam Vrankulj, covering the topic at Biometric Update, recalls that “The Electronic Privacy Information Center (EPIC) filed a FOIA lawsuit against the FBI in April to obtain documents related to the NGI.”

That’s a good catch, and it offers the opportunity to revisit the EPIC suit, assessed at the time here, in: EPIC sues FBI over biometrics FOIA request, where we noted EPIC’s tendency to overshoot the mark where technology is concerned.

The EFF requests are, to summarize, for:
1. Records related to the FBI’s proposed relationship with states to “build out its facial recognition database”
2. The FBI’s plans to combine civil and criminal data
3. Records related to the reliability of facial recognition capabilities

When compared to those of EPIC, the EFF FOIA activities certainly reflect a more moderate approach that would appear to have a higher likelihood of bearing fruit. The EFF seems carefully to avoid asking for information that the FBI can’t provide, and it would appear to be an easier request to comply with than EPIC’s.

UPDATE:
The EFF’s FOIA request/lawsuit has borne fruit.
FBI Plans to Have 52 Million Photos in its NGI Face Recognition Database by Next Year (EFF) Read te whole thing.

Face Recognition above the fold at the Washington Post

In the Sunday edition yesterday, the Washington Post ran a long piece above the fold on facial recognition, photo databases, and law enforcement.

State photo-ID databases become troves for police (Washington Post)

It looks like the issues we have been discussing are finally going mainstream.

UPDATE:
For fashion tips on how to beat facial recognition, check out CV Dazzle.

Compared to what?

Facial recognition software flops in US trials (Information Age)

“The number of system-generated false positives was excessive,” concluded the report, recently obtained under freedom of information regulations.

The airport had installed two separate facial recognition systems at security checkpoints at the airport. However, they failed to detect volunteers posing as terrorists 96 times during the three months that the trial was running, despite successfully picking them up 153 times.

Logan Airport was where 10 of the 19 terrorists involved in the 11 September terror attacks on New York boarded their flights.

Articles like this only make sense with reference to the expectations of the people that write them and users of the technology.

Catching a bad guy in 153 out of 249 chances (a 61.4% success rate) is obviously worth something, especially if the chance of catching that bad guy was 0% before the addition of the new technology. [Note: this analysis only makes sense if it is assumed that the subjects involved are on a watch list.]

The article also mentions the September 11 hijackers, begging the question: would the attacks have been possible if 11 or 12 (out of nineteen) of the attackers had been detected on the day?

Maybe, maybe not. Security protocols, not technology, would determine the answer. But a smart security protocol might say that if six terrorists are caught entering the same airport within a couple of hours of each other, certain measures should be taken.

False positives and low accuracy are real concerns to be overcome by improved performance of facial recognition systems. In fact, they are interchangeable problems. SecurLinx can provide a system that rarely provides false positives, but will fail to alert the user to a larger proportion of accurate matches, or we can provide a system that will catch more matches but will generate more false positives. We work with end users to help them determine their own “sweet spot” for the kinds of matches to which they want to be alerted. Facial recognition in surveillance applications isn’t like fingerprint biometrics.

The key to getting these things right for the customer and delivering the goods on a Return-on-Investment basis, is good communication about system capabilities, good training, and the application of the technology at the appropriate job function level.

If the system was sold as a bulletproof terrorist finder, a 39% failure rate is a flop. If it was sold as a 61% chance of preventing disaster, isn’t that worth something?

See also: Biometrics & ID infrastructure: Perfect is the enemy of good

India: Kerala police adopting face recognition for surveillance

Sree Padmanabhaswamy Temple
Photo: Rainer Haessner

Face-recognition tool to curb crime (The Hindu)

The State police will soon have the latest face-recognition technology integrated with its expanding surveillance camera network to screen entry and exit points of airports, railway stations, stadiums, and key government offices for persons with criminal or terror links. Senior police officers say the technology is likely to be implemented first at the landmark Sree Padmanabhaswamy temple here and later at other locations, including vital establishments vulnerable to sabotage in Kochi and Kozhikode.

“Any sufficiently advanced technology is indistinguishable from magic.”

— Arthur C. Clarke

Fingerprint led to arrest in Dollar General killings, detective testifies (Wichita Eagle)

Detective Tim Relph said a video camera showed the killer walking into the store and quickly leaving after shooting two people with a .22-caliber handgun. The killer tried to exit the store through an entrance door before realizing that the door wouldn’t open from the inside, Relph said. The finger and palm print left on that entrance door proved to be the key to solving the case, he said.

The shooting occurred at 8:01 p.m. on Nov. 30, Relph said, and a computerized fingerprint classification system identified Marshall as a possible suspect by 3:45 the next morning. By 4 a.m., he said, a fingerprint examiner confirmed that the print came from Marshall.

“By 4:30 in the morning there were 50 police officers looking for him,” Relph testified.

Shooting at 8:01 PM. Positive ID before 4:00 AM. That’s less than eight hours. Sometimes, we’re led by various television programs and movies to believe that the process is much quicker than that.

In actuality, given the steps involved, the eight hour turn-around is magical. Because…

Law enforcement, NIST making fingerprint files easier to search (GCN)

Not all AFIS are alike, however. State and local agencies often maintain their own databases, and although there can be some interoperability in a vertical hierarchy of local, state and federal databases, there is very little interoperability horizontally between neighboring jurisdictions. To search different databases, examiners must mark distinctive features for fingerprints manually for different systems, using different coding, notation methods and data definitions.

See also: Law enforcement interoperability, though little discussed, is a big deal

It looks like quite a lot of progress is being made on the interoperability challenges we’ve discussed from time-to-time.

US & EU to help Nigeria with fingerprint biometrics in counter-terrorism effort

Insecurity: US, EU renew support for Nigeria (The Nation)

“We have figure prints of possible over 10 million travellers at the same time in a system. We are expanding in Nigeria, Chad, Burkina Faso and we are doing a major upgrade in Ghana. We are possessing about 10 thousand finger prints per week in West Africa.”

When reporters sought to know what the US stood to gain in the partnership, Moro responded that the assistance was at no cost to the country but an extension of a hand of fellowship from a caring ally.

Other members of the delegation are: Mr. Dwight Brown, Miss Theresa Keens, Mr. David Svendsen, Mary Johnson, Thaddaeus Hoyt and Diana Kohn, who are programmes personnel at the US Embassy.

The European Union also renewed its continued support to the Federal Government “until terrorism is defeated”. Ambassador and Head of Delegation of the EU to Nigeria and the Economic Community of West African States (ECOWAS) Dr David MacRae, dropped the hint at a media luncheon hosted by the commission in Abuja.

This news article from yesterday provides important context.

Nigeria: Boko Haram Threat Chokes Trade With Cameroon (All Africa)

Cameroon has stepped up security over the Boko Haram (BH) threat.

In November 2011, Nigeria shut its border with Cameroon, prompting Yaoundé to bolster security in the largely Muslim Far North Region, close dozens of Koranic schools and hand over suspected BH members to Nigeria, which reopened the border in 2012.

Despite the intensified security, suspected BH militants on 19 February abducted seven French tourists, including four children, from a national park in the Far North Region, freeing them two months later.

Cross-border trade sustains the local economy in the Far North Region which sells onions, rice, maize, livestock and other agricultural goods to Nigeria, and imports sugar, cement, textile and electronics.

“Tight border security and checks are making business impossible for some of us. This was worsened by the kidnapping of [the French] tourists. Today all the goods must be checked before entry, and taxes are so high,” said Doudou Yaouba, a trader in Maroua, the regional capital.

This last, detailed, article illuminates major concerns within ECOWAS and among other interested parties. Biometrics can be a leap-frogging technology for providing domestic services, as in India’s UID project. They can also be a leap-frogging technology for bringing less rigorous international security protocols into a standard operating environment where highly sophisticated capabilities can be brought to bear.

Putting the mosaic together in Boston

The post’s title refers to the mosaic of information that can be arranged into a picture of the events leading up to the savage acts. The other mosaic, the way things were for so many unique individuals, can never be put back together.

How This Photo of the Boston Marathon Gives the FBI a Bounty of Data (Wired)

The photo — click to enlarge — shows a lot of people, what they’re wearing and where they’re positioned within the crush of Marathon fans. It’s important to law enforcement, as it “can be of use in putting the mosaic together,” says Robert McFadden, a former Navy terrorism investigator. Crabbe’s wide-angle panoramic photo “could be one of the many critical pieces of the map of the investigation.”

The panorama photo was one of seven shots Crabbe snapped with her phone during a leisurely stroll and later handed over to investigators.

The Wired article starts with a single data point (data set, really), a photo, and follows it part-way through the process the FBI has used during its investigation of the recent bombings in Boston.

…putting the mosaic together. It’s a good metaphor for how the people charged with figuring out what happened and who did it go about their work. Read the whole thing.

Also see:
What’s Going on Behind the Scenes of Bombing Investigation? Forensic Scientist, Former DHS Official Shed Light on Tech and Tactics (The Blaze)

“Facial recognition technology will play a very small part,” Schiro told TheBlaze in a phone interview.

“A lot depends on the quality of the images you have to work with,” Schiro continued noting that lighting, angle and other factors could really limit the use of facial recognition in the case. Not only that but there would need to be some sort of match for it to recognize.

UPDATE:
Here’s another good article about facial recognition and crime solving. I selected the two paragraphs below because they highlight both the organizational issue of interoperability and the technology issues around matching. There are other interesting insights in the rest of the piece.

Facial Recognition Tech: New Key to Crime Solving (The Fiscal Times)

However, it’s likely the FBI was unsuccessful in identifying the suspects using FR because either they didn’t have a quality image of the wanted persons, or the suspects were not in any of the databases the FBI has access too, Albers said.

While facial recognition technology has high-accuracy when used to match a clear image of a person with another passport-style photo, it is not as effective when used with low-quality images like the ones the FBI released on Thursday. The standard for facial recognition to be accurate requires 90 pixels of resolution between the two eyes of the pictured person. The pictures the FBI released of the suspects were about 12 pixels between the two eyes, said Jim Wayman, the director of the National Biometric Center.

and..
Facial-recognition technology to help track down criminals – Humans are still better at it (Kuwait Times)

Search for Boston bombers likely relied on eyes, not software (Reuters)

These last two reminded me of the (Facial Recognition vs Human) & (Facial Recognition + Human) post from November 2011.

In the Boston case, it looks like there were two barriers to effective use of facial recognition technology in identifying the suspects. On the “evidence” (probe) side, the image quality was poor. On the enrollment (database) side the only “correct” match was likely to be in a very large database such as the Massachusetts DMV database.

If only one of these conditions were true — for example a bad probe against a small database, or good probe against a large database — facial recognition technology might have been of more help.

Crowd-sourcing the ID challenge to a large number of human beings that operate with a lot more intelligence and information than facial recognition algorithms is another option. It’s been used with photographs since at least 1865 and without photographs since at least 1696.

One crowd-sourcing fact that law enforcement officials must consider, however, is that the suspect is almost certainly in the sourced crowd. If the suspect already knows he’s a suspect, that’s not a problem. If he doesn’t already know he’s suspected, that information is the price of getting the public’s help which means facial recognition technology will retain its place in the criminal ID toolkit.

UPDATE:
Boston police chief: facial recognition tech didn’t help find bombing suspects (Ars Technica)

“The technology came up empty even though both Tsarnaevs’ images exist in official databases: Dzhokhar had a Massachusetts driver’s license; the brothers had legally immigrated; and Tamerlan had been the subject of some FBI investigation,” the Post reported on Saturday.

Facial recognition systems can have limited utility when a grainy, low-resolution image captured at a distance from a cellphone camera or surveillance video is compared with a known, high-quality image. Meanwhile, the FBI is expected to release a large-scale facial recognition apparatus “next year for members of the Western Identification Network, a consortium of police agencies in California and eight other Western states,” according to the San Jose Mercury News.

Law enforcement interoperability, though little discussed, is a big deal

Tyneside jewellery heist could lead to DNA sharing (Chronicle Live)

A jewellery heist on Tyneside has sparked a review of DNA sharing across Europe that could force police to hand over criminal records to foreign counterparts.

Specialists in Newcastle will spearhead a £1.2m effort to design a database that profiles crimes committed across the continent as part of a controversial EU information sharing treaty.

It comes just 12 months after a convicted murderer and his armed gang from Eastern Europe were convicted of carrying out an armed raid at a Newcastle jewellers.

Led by convicted murderer Marek Viidemann, the ring was linked to at least 150 armed robberies across the UK and Europe before being eventually jailed for a total of more than 30 years.

First, DNA is likely to be a small part of whatever system improvements emerge.  It’s expensive and slow compared to just about any other biometric modality or combination of modalities such as finger, face and iris.

From a management standpoint it seems that if you want to have a free flow of people, you need to have a free flow of law enforcement information. This is easier said than done. It’s often a challenge even when dealing with adjacent counties in the same state in the US much less, as in the European context, two different countries.

The term for this system compatibility and ability to effectively cooperate among departments is interoperability. It is a managerial and technical challenge that is rarely dealt with in popular depictions of how law enforcement works but, especially as the complexity of the law enforcement challenge increases, it is of critical importance.

Often, there are good systems in place for passing information “up the chain of command,” i.e. from street cop all the way up to a state or national information repository, but the information doesn’t always flow as freely back down again in the other direction. For various reasons, the formal links between street-level law enforcement officers in neighboring jurisdictions run up through a centralized authority and then back down again, though there are often informal links that bypass the up-and-back-down information flow model. The implications for efficient multi-jurisdictional law enforcement are clear.

Some of these issues came up a couple of years ago in a post. Usefulness of Biometrics in Law Enforcement: Who is the Customer? The analysis there can be extended from biometrics to all sorts of law enforcement IT systems and it has a great deal of bearing on issues like the ones raised by the Newcastle jewelry heist by international criminals.

Many police professionals put a lot more into databases of all types than they ever get out of them. Through biometric technologies and other integration services, SecurLinx works hard to balance that out a bit for our law enforcement customers.

Facial recognition and criminal investigations

Updated & Bumped…
Facial recognition and identifying suspects in the Boston Marathon bombing (Biometric Update)

Since the bombing, authorities have appealed to anyone who was at the event to turn over any video or photo evidence, in hopes that they may hold crucial details in identifying a suspect. Though nothing can be confirmed, particularly in the midst of an investigation, it is likely facial recognition will be used to identify suspects from these photos and videos. So far, the FBI has received more than 2,000 tips.

Auhor Adam Vrankulj strikes the right tone in his article about how facial recognition is going to be applied to investigation of the recent terrible events in Boston. Read the whole thing.

Paul Scheupp, CEO of face recognition technology developer Animetrics weighs in on Fox’s “On the Record,” last night.

UPDATE:
Facial recognition technology could prove key in Boston Marathon bombing manhunt – video embed unavailable.

UPDATE II:
The Boston Marathon bomber: Caught on film? (BBC)
More personal videos are being shot now than ever before, and such footage could help identify the Boston marathon bomber. But how is that footage processed – and could civilians really solve the crime?

Philippines expands biometric law enforcement capabilities

Philippines: US donates two biometric machines to Immigration (Business Mirror)

Lawyer Maria Antonette Mangrobang, BI acting intelligence chief, said with the biometrics equipment the bureau will now to be able to build a wider and more reliable database of the illegal aliens and foreign fugitives wanted by immigration intelligence personnel.

He said, henceforth, the machine will be used to scan the fingerprints of all arrested aliens, and the data will be kept in a database along with their photographs.

TSA wants a slick new ID & document management system

TSA To Purchase Credential Authentication Technology In 2014 (HS Today)

A CAT system must verify the identity of air passengers and confirm they are able to travel beyond the security checkpoint for boarding their planes, TSA said in an announcement Monday. It must display authentication credentials to a transportation security officer (TSO) or other qualified operator and ensure proper ticketing of the passenger, instructing the system operator as to what action to take if the passenger is not yet cleared.

The CAT system also must integrate a credentials scanner, technology to authenticate credentials, a graphic user interface (GUI) and an application programming interface (API). Under a separate contract, TSA already has produced the API, which provides an interface with its Security Technology Integrated Program (STIP) for the transfer of passenger data.

Pretty cool. This sounds a lot like the IDTrac product we developed several years back, only we used IDTrac to keep track of bank checks instead of boarding passes. The facial recognition part of what it does is pretty elegant if I do say so myself.

UK: Immigration politics and biometrics

The United States isn’t the only place where immigration politics — and the role of biometrics — are coming to the fore. They’re very timely issues in the UK, as well.

WorkPermit.com covers Prime Minister David Cameron’s policy prescriptions like the dew covers Dixie, here: Cameron announces tough reforms to UK immigration.

The Guardian provides some analysis here: Immigrants’ residents permits: how would they work?

The repeated refusal of GPs, social housing officers and social security staff to act as immigration officers also means that if more robust residence tests are to be introduced for other EU nationals then an easy and authoritative way is needed of checking how long they have been in the country and what their immigration status is.

Ministers have confirmed that they are looking at plans to take fingerprints and other biometric data to be stored on a card with a photograph and electronic signature from new arrivals from next year.

It is within this context that the beleaguered UK Border Agency is being broken up (BBC). The UKBA is currently responsible for border protection, visa & passport issuance, asylum cases, immigration law enforcement, etc.