Increasingly powerful surveillance tools have shifted the power dynamics between the public and the spaces the inhabit.

Performance Surveillance seeks to redress this balance by adopting the tools of computer vision, artificial intelligence and machine learning using in ”smart” surveillance system to create a performance. Using live feeds sourced via unsecured CCTV cameras projected in a real time, the people captured in these feeds will become performers controlling what we see in a gallery space. As they move through the picture their gestures will be interpreted by Leon Butler´s machine learning model using the same methods used in surveillance ecosystem that allows for a real-time multi-person performance. The viewer will be able to cycle through different feeds which maybe be a busy town square or an empty beachfront, a station platform which dependant on the time of day where the camera is located with feed the interpretation of the picture. Butler will build layers of performance inviting artist local to co-opt the feed and adapt the system to their practice. Through these performances he will engage people to reimagine their relationship with surveillance, machine learning and technology to reimagine and better understand the world.