I created this installation with century old glass-plate negatives, face detection data gathered by Machine Learning, and handbuilt, unique sculptures.

I like exploring the discrepancy between how we apprehend each other as humans, and how Machine Learning algorithms allow computers to “see” and recognize us.

As found portraits come in and out of focus, a face detection visualization increases and decreases in accuracy. These video loops were created with Processing and face landmark data was captured with Runway ML.

The dimensional quality of the piece encourages visitors to approach the work and inspect it closely. As all 3D models are collections of triangles, the sculptures reflect the underlaying data of the visualiztions.

This work was conceived and sketched in 2019, but I didn’t see the finished piece until it was fully installed at Gray Area in San Francisco in 2021. It was wonderful to realize three frames and six generated portraits at once.

I hope to exhibit the piece again, and plan to add to the experience with new portraits and generative patterns.