machine learning / surveillance / moving image / performance

HD VIDEO, 2- Channel Video Projection, synchronous loop

Against the backdrop of the US military backed  data collection for advancing behavioral prediction models, the term coined by Harvard professor Shonshana Zuboff,  Watch it From Where They are surveys the undocumented immigrant experience of living under a state of surveillance control, the mechanism of which relentlessly contributes to the erasure, deportation, discrimination, and other forms of political violence upon the individual.To avoid surveillance and ICE, the immigrant bodies sought to become invisible, minimizing their appearances of any forms in fear of unwarranted media dissemination. This invisibility, however, as Hito Steyerl has pinpointed, sometimes does more if not equal harm than being visible in order to fight for the rights of legal representation in the long run.

Watch it From Where They are negotiates the seemingly incongruous relationship between invisibility and visibility for those whose data information is prone to the surveillance control. A set of bodysuit is made from machine generated adversarial patches that are used against popular object detection algorithm YOLO4 (open source code in courtesy of Simen Thys, Wiebe Van Ranst, Toon Goedemé). While the visually arresting patterns have made the subject into a public spectacle, it attacks the algorithm which then renders the subject invisible in the eyes of machine.

Juxtaposing the screens side by side with two clips playing simultaneously on the same time stamp, the two channel video installation invites the audience to watch the clips in the style of a real time surveillance duo monitor. As soon as the viewer is engaged in the activity of watching, they become complicit in the disappearing of the subject matter. Under the omnipresent gaze in both private and public sectors, the immigrant body is inevitably turned into a literal and symbolic data fed into the incessant collecting and classification of the illicit and racial topography of the Other in the digital landscape.

Director: Tracey Shi
Director of Photography: Dazhi Huang
Starring: En Tung Liu
Editor: Dazhi Huang
Photography: Caren Ma