Museum of Borderlands is an interactive installation that explores algorithmic decision-making in physical space. The work has been selected to CVPR 2021 Computer Vision Art Gallery. It’s my final project for MA Computational Arts.

The project invites us to explore the kind of visuals that become labelled as threats. What kind of new connections are born when images are increasingly grouped by their similar form in machine vision?

The work takes the form of an iterative museum that seeks to map our current fears. By inviting the visitors to reclassify its machine learning system, Museum of Borderlands offers the audience a place for building alternative classification models. The project is my first experiment in gathering collective, subjective datasets for deep learning.

The installation is driven from a web app that classifies the webcam feed with Tensorlow.js and sends OSC messages to trigger audio in MaxMSP.