IS IT ME? PORTRAITURE IN THE AGE OF FACIAL RECOGNITION

 
                                                      Is it me? oil on linen 51 x 76 cm 2020

I started this self-portrait - or is it? - earlier this year. Although I thought it was finished I have worked on it and have just finished it. 

As regular readers know, during the pandemic I have been responding to media coverage about the use of drones, and problems with creeping normalisation of drone use.* Uses have included monitoring population compliance with COVID19 restrictions, broadcasting instructions and spraying disinfectant. Other uses that have been mooted, but I am unsure if they have actually been carried out, are checking for temperatures and heart beats to assist determination of illness. Issues of facial recognition have escalated over the last few years, and the pandemic, with co-current protests around the world about various issues, has highlighted issues of privacy, AI accuracy, bias and security.

Deep Fake and GAN Technology
While Is it me? intersects with the various issues of facial recognition, drone use and contemporary surveillance, it also triggers questions about portraiture in the age of AI and machine learning. Facial recognition technology uses biometrics scoped from photographs or image data to map facial features. These features are tagged as reference indicators for identity recognition by AI systems. Deep fake technology and generative adversarial networks (GANs) produce images of people. The latter has been used to generate portraits - note I use the word generate, rather than create. A recent controversial example of GAN 'portraiture' is the auction of the Portrait of Edmund Belamy. Sold for $432,500, the image of Belamy, a fictitious person, was generated and printed after the GAN system had been fed a "data set of 15,000 portraits painted between the 14th and 20th centuries". Even thinking about portraiture across centuries being  called a "data set" is unnerving! 

Portrait Painting

With technology changing the ways we identify and portray people, what can portrait painting say?  Is it me? is my attempt to grapple with that question. For me, my two main characteristics are my very blue eyes and my bun at the back of my head. In Is it me? I have featured these two characteristics as dominant, as if they are standing up to the computer graphic biometric indicators that scan my face. With so many images in the media of these kinds of computer graphic markings scanning faces, my inclusion of them is a parody, but also a warning. Like our natural features, will these algorithmically determined indicators become part of our 'portrait', who we are? Historically, though, portrait painting has been far more than simple identity. Good portraits are ones that reveal a personality, a sense of spirit - whether it is joyous or not. Good portraiture can also tell a story about a person. This is often achieved by combined elements such as, whether their skin is rough or smooth, what they might be wearing, the background against which they are portrayed and expression. With abstract approaches to portraiture, recognition and story-telling are bound up in painterly gesture, placement of forms, use of colour. 

Biometric scanning, gait monitoring, and surveillance of online activity and physical movements detected through devices, can be combined to 'tell a story' too. But, the purpose of the story is often linked to tracking and targeting. The 'story' becomes data, accessible for further tracking, surveillance and targeting by civilian, corporate and government entities. In the case of military, terrorist, counter-terrorist or criminal activities, tracking and targeting take on life and death questions. The point here, is that stories imparted by more traditional notions of portraiture are more ambient, and normally less instrumentally invoked.

           

ERROR?

In Is it me? I have also included a civilian drone. Its sensors are obviously scanning my face. This is a direct response to concerns about creeping normalisation of drones used for monitoring/surveillance during the pandemic period. The red ERROR suggests this portrait, according to the facial recognition technology, is not me! A small question mark after ERROR seems to indicate that the ERROR message is wrong. The smallness of the question mark channels a sense of confusion, as if the surveillance system is not sure. The flat horizon of the landscape against which I am painted, is a reference to my story. It is the landscape of my childhood, the flat, naturally treeless, black soil Pirrinuan Plain, outside Dalby on the Darling Downs, Queensland, Australia. This landscape reference is not just about where I grew up, it is also about me being an artist. The vast landscape of my childhood, where distance was my challenging playground, is still the launching pad for my imaginational flights into the cosmos. It is the foundation for my thoughts about imaginational metaveillance  

There is a lot more to say, but I will leave that up to you to think about! 

                                    Me in 2018 on a visit back to my childhood landscape

* Other posts about creeping normalisation of drone use during the pandemic

Dronescape: A Creeping Normalisation 
Border Crossings 
Watercolours for our Strange Times                                   

NEWS

My submission to present Drones, Art and Risk Analysis at the 2020 World of Drones and Robotics Congress, in Brisbane, in November, has been accepted. I will be presenting with a short film of me talking, rather than in-person. 

Cheers,

Kathryn