Terence Broadمشاهده پروفایل
پژوهشگر
Terence Broad is a researcher at Goldsmiths, University of London , specializing in generative neural networks and computational creativity. His work bridges machine learning with artistic expression, focusing on techniques like network bending and active divergence to push generative models beyond data imitation. Broad's research includes data-free methods, divergent fine-tuning, and expressive manipulation of deep generative models. Thesis : Expanding the Generative Space (2025) presents novel approaches for active divergence in generative systems. Key Contributions : Pioneered network bending frameworks for feature manipulation, explored uncanny amplification in deepfakes, and developed autoencoder-based film reconstruction systems. His publications span conferences like ICCC’21, EvoMUSART 2021, and SIGGRAPH 2016, demonstrating applications in image, audio, and video domains. Broad's collaborations with Frederic Fol Leymarie and Mick Grierson highlight interdisciplinary approaches to machine-generated creativity.