Confronted. 2018. with Tatyana Mustakos and Char Stiles.
This project uses GANs to allow the user to draw a being into existence. It's meant to be a small window into the magic of bringing something new into the world and facing your creation.
Listening. 2018. with Jeena Yin, Joey Gibli, and Anirudh Mani. This project uses RNNs to mimic the experience of eaves dropping on a 2 way conversation. The viewer comes closer only to realize the conversation isn't as predictable as it seems.
These images are artifacts from the data analysis I collected on the Teenie Harris Archive from the Carnegie Museum of Art. This archive is one of the most comprehensive visual datasets of African American life in this country.
Using edge detection, regression, clustering, and neural networks I was able to gather statistics such as the age and gender distribution of the subjects as well as different compositional elements that are key to Teenie Harris' photos.
I applied techniques like object detection, depth prediction, and saliency prediction to provide a deeper understanding of this archive of 60000 photos of african-american life.