Can facial recognition assist art historians?

Art research effort aided by face recognition (BBC)

The software under development by Prof Rudolph, engineer Amit Roy-Chowdhury and art historian Jeanette Kohl will try to put a name to these forgotten likenesses by grabbing data about defining elements of faces from portraits and comparing them to known depictions.

Early work on the project established that key parameters for facial recognition in portraiture include the position of the corners of the eyes and mouth, the width between the eyes, and the width of the mouth. Mapping these characteristics using a 27-point scheme captures sufficient information to make identifications, said the researchers.

Benjamin Franklin death mask.
Source: UndyingFaces.com

The application of facial recognition technologies to the challenges of art history is a fascinating topic. We once had a call from a collector who believed he had discovered a third existing photograph of Henry David Thoreau and wanted our help with the verification.

The BBC article, however, discusses the application of facial recognition technology to artifacts created before the invention of photography, which presents a whole different set of challenges.

Art historians interested in applying software to facial analysis might be wise to enroll death masks (also mentioned in the article, or life masks) as database images since they are as nearly as possible an actual recording of a person's face. Those could then be compared to painted images purportedly of the same person. Using this technique, art historians would quickly get an idea of how "accurate" portraitists (in general, or individually) were at painting images to the mathematical specificity that underlies facial recognition algorithms.

They would probably learn a lot from this basic exercise. How accurate did portraits tend to be? Were they supposed to be scientifically accurate? Or, were they more like the airbrushed and Photoshopped images that grace the covers of magazines today? Were certain portraitists consistently more accurate than others. Once these questions are well understood, analysis of painted or carved images for identification purposes might come within reach even where no contemporary physical impression of the face exists.

Perspecitves on ID in earlier- vs. later-developing countries

The Government and the UK's National Technical Authority on Information Assurance (CESG) have published new guidance on 'identity proofing' and verification. (Pinsent Masons)

"Within the UK there is no official or statutory attribute or set of attributes that are used to uniquely identify individuals across Government," the joint Cabinet Office and CESG guidance document said. "Neither is there a single official or statutory issued document whose primary purpose is that of identifying an individual. Without such attributes or documentation it is difficult for any person to be absolutely certain of the identity of another."

"This guide is designed to demonstrate how a combination of the breadth of evidence provided, the strength of the evidence itself, the validation and verification processes conducted and a history of activity can provide various levels of assurance around the legitimacy of an identity," it said.

The whole piece is interesting.

The first quoted sentence above really jumps out, though.

The early industrializers/bureaucratizers typically developed their ID schemes in an ad hoc fashion. The church kept its records for its purposes. The military kept its records for its purposes. Schools, for theirs. Service providers, etc. The system generally works. In the end, error rates and whether or not the costs of the ID errors exceed what it would cost to fix them rule the day. Political and financial considerations factor in.

It is precisely this patchwork ID environment that later-developing countries are choosing to leap-frog with more centralized (United Arab Emirates) or ecosystem (India) approaches involving biometrics. Outside observers from the earlier developing countries are often surprised that their political perspective on government-backed ID isn't universally shared while observers in later-developing countries may be equally surprised that the most developed countries in the world have such patchwork ID systems.

Interview with James Wayman

Very good interview at Institute of Electrical and Electronics Engineers (IEEE) with James Wayman on face recognition and other biometrics. Mr. Wayman is the former director of the National Biometric Test Center at San Jose State University and is now an administrator in its Office of Graduate Studies and Research.

Will Face Recognition Ever Capture Criminals? (IEEE.org)

Okay, so, let’s start historically with the technology. It was developed in the early 1960s by a fellow named Woodrow W. Bledsoe, who I believe was an IEEE member. He later retired at the University of Texas at Austin. And what he was doing was marking facial images by hand—the centers of the eyes, the corners of the eyes, the corners of the lips, and the like. And then he projected these marks onto a sphere and he rotated the sphere, trying to get marks from two different images to line up, at which point he could say, aha, these are from the same person.

Well, all of this hand marking didn’t work so well, and in the late 1980s, Sirovich and Kirby came out with this very simplistic idea that is so simple it sounds like it’ll never possibly work, but it did."

For those who prefer listening to reading, there's also a podcast...

The past and future of the American passport

The existence of the book, The Passport in America:The History of a Document (Amazon) somehow escaped my attention.

Luckily, the Boston Globe ran a short Q&A with the book's author, Craig Robertson, today. Click on over to see his answers to the questions below.

When were Americans first required to carry passports for foreign travel?

What was the purpose of passports issued by the State Department before World War I?

Nineteenth-century passports actually featured written descriptions of their holders, correct?

So what was the reaction among the American public when photographs were introduced to passports?

In the days before birth certificates became commonplace, how did the government verify an applicant’s identity?

What future changes to the passport do you foresee?

In your research, you probably looked at hundreds of passports. Did you find a single good passport photo?

The Crime Wave of 1920 and the Making of the Modern FBI

Before the dawn of the Twentieth Century, The Bertillon System was the standard for biometric identification.

In 1903 the New York state prison system had begun to use fingerprints. By 1908, all the branches of the U.S. military had adopted fingerprints. In 1924, an act of congress established the Identification Division of the FBI. The IACP's National Bureau of Criminal Identification and the US Justice Department's Bureau of Criminal Identification consolidated to form the nucleus of the FBI fingerprint files (source - and a very interesting site in its own right).

How did we get there? Read on.

The Wall Street Bombing That Made Hoover and the FBI (Bloomberg)

Shortly after noon on Thursday, Sept. 16, 1920, a powerful bomb hidden in a horse-drawn wagon exploded at the corner of Wall and Broad Streets in Manhattan. It was a pleasant late-summer day, and throngs of people had been out enjoying a lunchtime stroll, a brief respite from the great money machine, the center of American capitalism.

Now blood ran in the streets where the first U.S. Congress had convened and the Bill of Rights became law. Shrapnel scarred the walls and shattered the windows of J.P. Morgan and Co., America’s most formidable bank. The bomb killed at least 38 people and injured roughly 400. It was the deadliest terrorist attack in U.S. history, a distinction it held for 75 years. Its force reverberates today.

In Washington at that hour, J. Edgar Hoover, 25 years old, was putting the finishing touches on the federal government’s first counterterrorist force, the General Intelligence Division. Hoover wrote that he intended to combat “not only the radical activities in the United States” but also those “of an international nature”; not only radical politics, but “economic and industrial disturbances” as well.

...

Walk to the corner of Wall and Broad Streets today, and you can run your hands over the deep gouges left by the 1920 bombing. You will have to look harder to see the cameras that track your steps -- a 21st-century tribute to Hoover, the architect of the modern surveillance state. Every fingerprint on file, every byte of biographic and biometric data in the computer banks of the government, owes its origins to him.

See also:
The Bertillon System: An Early ID Management System
The History of Fingerprints (and the Death of the Bertillon System)

The History of Fingerprints (and the Death of the Bertillon System)

Following up on the previous post on the Bertillon system, the Bertillon system was supplanted by fingerprints.

This page at onin.com has an excellent run-down of the history of the recognition (and use) of fingerprints as unique to each individual.

Remarkably, the assumption that fingerprints are, in fact, unique dates back to the dawn of history.

The linked page also has an account of the wild circumstances surrounding a famous mis-identification in 1903 that is credited with precipitating the demise of the Bertillon system as the primary identification system used for purposes of law enforcement.

The Bertillon System was generally accepted for thirty years. But it never recovered from the events of 1903, when a man named Will West was sentenced to the U.S. Penitentiary at Leavenworth, Kansas. It was discovered that there was already a prisoner at the penitentiary at the time, whose Bertillon measurements were nearly the same, and his name was William West.

Upon investigation, there were indeed two men who looked exactly alike. Their names were Will and William West respectively. Their Bertillon measurements were close enough to identify them as the same person. However, a fingerprint comparison quickly and correctly identified them as two different people. (Per prison records discovered later, the West men were apparently identical twin brothers and each had a record of correspondence with the same immediate family relatives.)

Criminal twin brothers with the same name killed the Bertillon system.

UPDATE:
Thanks to our Chief Technology Officer, Steven, we have found photos of the infamous West brothers on the web site of The National Law Enforcement Officers Memorial.

Source: The National Law Enforcement Officers Memorial Fund; Photo

Here's a close-up of the picture above that fills in some detail:

The Bertillon System: An Early ID Management System

In a post the other day, I made a brief mention of the Bertillion system, also known as anthropometry, Bertillonage and the Bertillon system. "Bertillon system" seems to be the most common usage. Developed for use in criminal identification, the Bertillon System was an extremely important early attempt at using objective, measurable details of the human body for use in establishing an individual identity with a high degree of certainty.

Before 1882 establishing an individual's unique identity usually involved the testimony of trusted individuals. In a criminal trial, witnesses to a crime would swear that they did or didn't see the suspect commit a crime. Establishing a criminal history might rely upon a police officer under oath to establish the suspect's criminal record* [1].

With the invention of anthropometry, a system of body measurements of adult individuals for personal identification, Alphonse Bertillon, then working with the police in Paris, changed all that.

Divided into three integrated parts, Bertillon's anthropometrical system consisted of:

♦ Bodily measurements conducted with the utmost precision and under carefully prescribed conditions of a series of the most characteristic dimensions of bony parts of the human anatomy;

♦ The morphological description of the appearance and shape of the body and its measured parts as they related to movements "and even the most characteristic mental and moral qualities"; and

♦ A description of peculiar marks observed on the "surface of the body, resulting from disease, accident, deformity or artificial disfigurement, such as moles, warts, scars, tattooings, etc."[2]

Bertillon's system, which made the jump across the Atlantic in 1887, was greatly enhanced by advances in photography.

Here's a great example of a Bertillon record from Jersey City, New Jersey 1898 [3]

For a much larger view, Click here.

The Bertillon system was eventually abandoned world-wide because it failed to provide reliable and unique measurements, was too cumbersome to administer in a uniform manner and (unlike fingerprints) it didn't rely on a single measurement of any part of the body for identifying a specific individual.[2]

* Definitively establishing a criminal history that would merit harsher punishment than that meted out to first time offenders might (in certain times and places) also rely upon the highly effective, though irreversible, methods of dismemberment (cutting off the hands of thieves, etc.), branding, scarring and tattooing [1].

UPDATE: 
An observation: the measurements on the left half of the front of the record (the Bertillon measurements) appear to be in centimeters (metric system) while the height and weight on the right side are in standard units.

UPDATE II:
Something of a continuation of this post: The Death of the Bertillon system and the History of Fingerprints

Sources:
[1] Origins of the New York State Bureau of Identification by Michael Harling
[2] Alphonse Bertillon and Ear Prints by forensic-evidence.com
[3] Jersey City Police Department Bureau of Criminal Identification (B.C.I.), New Jersey, USA

See Also:
Alphonse Bertillon - Wikipedia.com

Fingerprinting in the workplace

Why fingerprints anyway? (Examiner)

Before the mid 1800s, law enforcement officers with extraordinary visual memories, so called “camera eyes”, identified previously arrested offenders by sight. Photography lessened the burden on memory but was not the final answer, since personal appearances change.

There's more on the history of ID, the Bertillion system, and current legal and corporate applications.