Facial Recognition Technology
There is hope to use FRT like a fingerprint (p. 17), but there is no universal way of seeing or of seeing the face (p.11). Photographs and archival processes have made us falsely believe in the unity of the face (p. 23). In FRT, many recognitions of face are specific to the use of the system (p. 6), and human ways of seeing faces change with society. "Face perception does not reduce to a universal psychological process" (p. 11-2). They hope to re-embody dis-embodied bodies (p.12) (i.e., giving a face to a data double). The technology promises to standardize and steady ambiguities of identity (p. 14) and deliver this information through computer networks (p. 15). Disadvantages are that faces are not stable (p. 17), but advantages are that these images can be taken without consent (p. 16). "Facial recognition systems promised to enable more effective institional and administrative forms of identification, social classification, and control" (Gates, 2011, p. 28).
TYPES
Different types of software are FRT and facial expression analysis. While FRT does not pretend to know what is going on inside someone, automated facial expression technology (AFET) does suggest it knows.
HISTORY
Galton spent time researching and classifying human features (although his view was clouded by racism) (p. 19).
PROBLEMS
1) "Identity" is dependent on built in algorithm designs, and the determination is thus limited by technological design constraints (p. 20).
2) Facial recognition is tied to social assumptions. Technology is not neutral (p. 21). Thinking that the face can be cut off from its "social, historical, and material context" as well as the body is problematic, and this type of thinking occurs when we assume that the face is the site of our identity and subjectivity (p. 23).
CCTV
Pairing FRT with CCTV equipment would create smart cameras that would serve as "interpassive surveillance" ("whereas interactivity implies a spectator-user actively engage with electronic media and taking part in the production of content and meaning, interpassive arrangements allow the medium itself to do the work of receoption for the user", p.73) which means that FRT software would perform the mental labor of watching and supposedly "bring a measure of objectivity to surveillance processes" (Gates, 2011, p. 74).
PHOTOGRAPHY
Photography "is now used a means of constructing, facilitating, and enhancing these dimensions of what it means to human. Of course it is also used to classify and individuate us in ways which over which we have no control, and these different impulses of photography -- [to quote Allan Sekula] its honorific and repressive tendencies" (p .23). "The circulation of facial images as indexes of identity occurred with the more general formation of what Craig Robertson has called a "documentary regime of verification.' an ad hoc process whereby techniques for administering standardized identification documents and using those documents to very identities came to replace local practices of trust and face-to-face recognition" (p. 33). The database is the crux of the application of photography and crime prevention (p. 102), and the "central artifact of criminal identification was not the camera, according to Sekula, but the filing cabinet" (p. 111). Archives establish order and create an idea of empirical truth (p. 111).
"The persistent need of the police for more effective means of organizing photos and identifying faces explains why the U.S. National Institute of Justice provided nearly as much funding for research and development of the technology as the Department of Defense in the 1990s" (Gates, 2011, 54). Law enforcement agencies teamed with computer corporations to help them develop mug shot documentation systems (Gates, 2011, p. 71). In 1995, the FBI hosted a "Mug Shot and Facial Image Standards Conference" (p 71).
FACES
According to Deleuze and Guattari, "all faces envelope an unknown, unexplored landscape" (qtd. in Gates p. 23).
AUTOMATED FACIAL EXPRESSION ANALYSIS (AFEA)
FRT fails to read what is inside a person and only focuses on the exterior (p 152), and "programmong computers to identify facial expressions, especially under uncontrolled conditions, poses a difficult set of problems" (p.153).
DRAWBACKS
FRT is not a natural process; organizations have to decide to use this technology. Organizations need to think of the complexity, cost and manpower; these are large systems which have many components and differ based on pricing and abilities. They must consider 1) that they will have to build and maintain databases; 2) establish policies, procedures and training for the technology and those involved; 3) how to deal with false positives; and 4) consider the technology may become obsolete with time (Gates, 2011, p. 58).
References:
Gates, K A. (2011). Our biometric future: Facial recognition technology and the culture of surveillance. New York: New York UP.