Uncanny Valley

Welcome to the world of Uncanny Valley—a whimsical exploration of facial recognition, social media, and augmented reality. I delved into tens of thousands of Snapchat selfies, unleashing the power of my GAN to create its own unique renditions. Surprisingly, the GAN often chose to discard Snapchat augmentations, resulting in a captivating array of unadorned selfies. I playfully re-introduced these into Snapchat, allowing anyone to overlay them onto their own faces. It’s a reflection on the multiple personas we craft on social media, an opportunity for some laughs, and a chance to share both silly and eerie selfies.

Picture a face manipulated in Snapchat, adorned with a computer-generated, melty mask—set against an artful background. My fascination with Snapchat led me to scrape images from Instagram, curious about what a neural network could create. The Generative Adversarial Network, instead of blending Augmented Reality components into nightmarish forms, astoundingly detected commonalities and erased these components altogether!

a selfie of me with a mask on and a segmented mode collapse background. it is mildly terrifying

Enter the world of Generative Adversarial Network grid at mode collapse. Working with a DCGAN until it reached mode collapse provided intriguing results, with converging grids making for visually captivating works. During training, I experimented with Snapchat’s face picker, leading to both horrifying and hilarious outcomes. The process unfolded in a detailed Twitter thread as I navigated the intricacies of Snapchat’s features.

a grid of GAN generated faces. they are deliberately janky

Moving forward, I employed style transfer on some images for upsampling and utilized resizing GANs and upsampling algorithms to ensure the works reached print size from their initial 128px.