A screenshot of the iPhone camera interface showing hanging masks resembling human faces in an art gallery, and yellow squares around each mask indicating that the phone camera has recognised a human face.

Did you know you can generate a portrait of a person’s face based on a sample of their DNA? The thing is, despite companies selling this service to the police to help them identify suspects, it’s not really that accurate. That lack of precision is at the heart of Heather Dewey-Hagborg’s work Probably Chelsea, a display of 30 masks showing 30 possible portraits of Chelsea Manning based on a sample of her DNA that she mailed to the artist from prison. The work is showing at Kunsthall 3.14 here in Bergen until the end of September.

Many masks resembling human faces hang from the ceiling in an art gallery.

DNA fingerprinting is when a DNA sample is compared to the DNA of a known individual to see if they match. That is fairly uncontroversial. Forensic DNA phenotyping is analysing a DNA sample to find out something about an unknown individual. Gender is easy. Geographical ancestry (20% Asian, 60% European, 20% Native American for instance) and externally visible characteristics (EVCs) such as eye and hair colour – but have a significant margin for error. DNA research has focused more on disease markers than on appearance, so this is not yet well-developed. Perhaps for good reason. Right now, we can predict pigmentation (so skin colour, eye colour, hair colour) better than other externally visible characteristics like the shape of a nose or arch of an eyebrow (Kayser 2015). Yet portraits generated from DNA samples are often presented as very accurate, as on the company Parabon’s website:

Parabon — with funding support from the US Department of Defense (DoD) — developed the Snapshot Forensic DNA Phenotyping System, which accurately predicts genetic ancestry, eye color, hair color, skin color, freckling, and face shape in individuals from any ethnic background, even individuals with mixed ancestry.

The thing is, these characteristics are then mapped onto a 3D model of the face. That can be quite accurate if they have access to the individual’s skull, as they might for the victim in a murder case, or where historians want to find out what a historical person looked like. If they don’t already know the basic structure of the face and skull, they use average faces based on the ethnic breakdown of the individual’s DNA. Parabon shows the process for mapping the general facial attributes to a specific skeletal structure here. As you can see the end result is fairly different from the starting point.

a series of five male faces marked 1 to 5 with the following text: Stages of a Snapshot Facial Reconstruction: (1) a Snapshot composite produced from DNA extracted from the subject's bone; (2) Snapshot composite with skull overlay; (3) Snapshot composite after rescaling to conform to skull dimensions; (4) a cutaway image illustrating near final composite; and (5) final, blended Snapshot composite.

Heather Dewey-Hagborg writes in her essay for the exhibition catalogue that “Probably Chelsea shows just how many ways your DNA can be interpreted as data, and how subjective the act of reading DNA really is.”A mask of a photorealistic human face is hanging in the foreground. Other masks hang behind it.

One of the striking things about visiting the exhibition with other people present rather than just seeing videos and photos of the silent masks in the online documentation is that you see how people are drawn to standing behind the masks. At the opening in Bergen I saw dozens of people taking photos standing among the masks. Here are two women examining the photos they have just taken.

Two women stand among hanging masks, looking at a camera.

I felt the urge too! Here I am:

A woman with brown hair stands behind a hanging mask of a human face.

Masks are of course a symbol of identity, and of identity as something that can be enacted, exchanged, used in a performance, put on and taken off. There’s something more, too, though: by imagining ourselves wearing these masks, we imagine ourselves stepping into Chelsea Manning’s possible identity. As Dewey-Hagborg writes, “We have so much more in common genetically than difference. Probably Chelsea evokes a kind of DNA solidarity; on a molecular level we are all Chelsea Manning.”

Of course, this artwork speaks directly to my ideas of machine vision. They are generated from data, both in the sense that DNA is algorithmically interpreted to create facial images, and in the sense of the masks being 3D printed. The computer-generated nature of these faces becomes clear if you view the masks really close. A close-up photograph of a single mask.

The faces are also computer-readable. See what my phone camera does when I try to take a photo of the masks? It sees them as human faces, human identities. It tries to set the focus on each of the faces, and once the photo is taken, it will try to match them up to the people it has already identified on my phone.

A screenshot of the iPhone camera interface showing hanging masks resembling human faces in an art gallery, and yellow squares around each mask indicating that the phone camera has recognised a human face.

I found it unexpectedly moving to walk into the gallery and be faced by this crowd of masks, silently staring at me. It is an intriguing experience. Reading about the ambivalence of DNA forensics makes me all the more intrigued.

1 Comment

  1. Suzanne

    That last photo showing you next to one of the masks shows that YOU could be Chelsea too!
    I think Heather’s point would be even stronger if she had generated masks for say 10 other people based on identical technology. Testing would have to be double blind, where both artist, DNA company and the technicians performing the testing do not know either the names, the gender, the ethnicity or anything else about the test persons. Actually it would be fun to throw in a few non-human primates to see if the DNA labs managed to pick up on species.

Leave a Reply to Suzanne Cancel reply

Recommended Posts

Triple book talk: Watch James Dobson, Jussi Parikka and me discuss our 2023 books

Thanks to everyone who came to the triple book talk of three recent books on machine vision by James Dobson, Jussi Parikka and me, and thanks for excellent questions. Several people have emailed to asked if we recorded it, and yes we did! Here you go! James and Jussi’s books […]

Image on a black background of a human hand holding a graphic showing the word AI with a blue circuit board pattern inside surrounded by blurred blue and yellow dots and a concentric circular blue design.
AI and algorithmic culture Machine Vision

Four visual registers for imaginaries of machine vision

I’m thrilled to announce another publication from our European Research Council (ERC)-funded research project on Machine Vision: Gabriele de Setaand Anya Shchetvina‘s paper analysing how Chinese AI companies visually present machine vision technologies. They find that the Chinese machine vision imaginary is global, blue and competitive.  De Seta, Gabriele, and Anya Shchetvina. “Imagining Machine […]

Do people flock to talks about ChatGPT because they are scared?

Whenever I give talks about ChatGPT and LLMs, whether to ninth graders, businesses or journalists, I meet people who are hungry for information, who really want to understand this new technology. I’ve interpreted this as interest and a need to understand – but yesterday, Eirik Solheim said that every time […]