[2022]
Win Back Damascus is a real-time generative imaging system designed and engineered by Rhea Dally. Built from a fragmented dataset of approximately 700 heterogeneous images, the system uses StyleGAN2 to generate continuous visual output driven by live audio input. Dally developed the full machine-learning pipeline, real-time audio-to-latent control architecture, and deployment as both a browser-based and standalone application. The work was presented in live performance and exhibition settings.
Related Work