MFA Thesis

 

Abstract


Gaze Relations is an art installation that visualizes how human gaze and computer vision see bodies in visual media to compare how they operate perceptually and culturally. Gaze, in visual media, refers to the way the filmmaker and audience look at bodies and how the filmmaker behind the camera influences the way the photographic subject is represented to an audience. Body detection, on the other hand, is a computer vision technique employed by surveillance institutions to automate the process of detecting bodies in surveillance footage. Gaze Relations critically visualizes and materializes these abstract phenomena of gaze and surveillance through visual art that represent the way these modes of vision operate. Gaze-tracking software tracks how a person looks at an image and then custom-built software reconstructs the image through their gaze demonstrating how they visually engage with the photographic body. With respect to surveillance, critical software was developed to visually analyze the processes of how a popular body detection algorithm models a human body. In addition to video, costume designs represent how the algorithm sees the body projecting the machine vision back into the world it is looking at. The piece highlights the differences between the complexities and nuances of human gaze to the algorithmic processes of computer vision that reduce the human form to simple features. Gaze and computer vision will be discussed as powerful social and cultural force capable of influencing the external world, not just the phenomena of taking in light and processing it.









MFA Show




MFA Thesis Paper

Thesis Paper Download