On Halloween, students taking Apryl Williams’ DIGITAL 359 course (“Gendered and Racialized Bodies and Technologies”) showed up to class over Zoom in elaborate makeup: glitter, face paint, hairstyles. But students weren’t in just any makeup—they were in CV Dazzle.
Based on a type of WWI British naval camouflage strategy, “CV Dazzle” was coined by artist Adam Harvey, whose project Computer Vision Dazzle (2010) explores how saturated colors painted in Cubist-inspired designs on faces and avant-garde hairstyling can trick facial-detection technology. Facial-recognition algorithms operate by identifying spatial relationships of key facial features, such as the nose, eyes, and mouth. In breaking apart the “continuity of a face,” those donning CV Dazzle can block detection by creating an “anti-face.”
After reading from Allissa Richardson’s Bearing Witness While Black and Simone Browne’s Dark Matters, students came to class ready to discuss how people of color can exercise agency to resist surveillance. “CV Dazzle is a way to avoid or disrupt surveillance, allowing folks to participate in civil protest,” explains Williams. CV Dazzle became especially important during the 2020 protests in support of Black Lives Matters, particularly because facial recognition technologies are more likely to wrongly identify the faces of Black individuals (which in the US has led to multiple wrongful arrests, including two in Metro Detroit). “I thought it would be interesting to see if the class could collectively trick Zoom’s vision.”
Together, the class played with Zoom’s built-in filters to see if Zoom’s facial recognition features worked when people were in CV Dazzle. (Williams intentionally planned this discussion for Halloween so that students would be more comfortable getting into the activity). Students who concealed their eyes had the most success: filters that applied animal faces or pirate faces to users couldn’t accurately locate the face if eyes were obscured. “That means Zoom’s computer vision is fairly adept,” notes Williams, “as we’d expect, since they have become a global provider of video communication.”
Hear it from the DIGITAL 359 students themselves:“I have never heard of CV dazzle before so it was interesting to see the different ideas and to create one of my own for our class. It is scary to think that a facial recognition system only needs to see a fraction of one of the facial key points to identify where other parts of the face are. [...] It makes me wonder if these platforms, especially snap chat, that opens the camera as soon as you enter into the app, is surveilling you even if you don't take a photo or if you’re on the camera for just one moment.” – Isabella Aarin Greenman
“I thought it was fascinating that the theory behind its invention was taken from a British military camouflage tactic. This fact alone illustrates just how much of a war-like country we are, that we feel obligated to put on camouflage just to express our beliefs publicly. Regardless of whether we have the right to protest, the state will always keep tabs on whoever they deem to be too non compliant and will take action accordingly.” – Drew Altman
“It’s interesting to see the approaches to distorting one’s face enough so it becomes unrecognizable, especially since it’s becoming increasingly harder to know what exactly the algorithm is seeing. [...] Additionally, I feel like it’s important to understand that algorithms aren’t seeing faces in the same way we do, so I feel like trying to outsmart them may be more of an artistic statement than anything.” – Mya Strayer
“It was shocking to me how easy it is to fool the technology, especially for people of color. We think that technology has a chokehold on us, but in reality, we are the ones that created it so we can figure out how to trick it.” – Ashley Krauthamer