Hubert Kompanowski
PhD Researcher in 3D Generative AI
About Me
I’m a researcher and engineer specializing in 3D Generative AI. Currently, I’m pursuing my PhD at Trinity College Dublin, where I focus on improving how AI handles 3D shapes, textures, and styles, especially making these models more controllable and reliable.
I have a strong background in software engineering and robotics. Before returning to academia, I worked as an ML and Software Engineer at companies like MIM Solutions and IBM. My experience ranges from building medical imaging tools to developing computer vision for Mars rovers, traffic surveillance, and other real-world systems.
I’m motivated by practical innovation: I’m co-inventor on eight U.S. patents (with IBM) and collaborate with startups to turn research into products. When I’m not writing papers for conferences like 3DV 2025, I experiment with AI filmmaking workflows to see how far we can push creative tools.
Selected Research
FROMAT: Multiview Material Appearance Transfer via Few-Shot Self-Attention Adaptation
FROMAT is a lightweight appearance transfer technique for multiview diffusion models. We use self-attention adaptation for high-fidelity material transfer from single reference image.
Dream-in-Style: Text-to-3D Generation using Stylized Score Distillation
Dream-in-Style is a method for stylizing text-to-3D generation. We introduce stylized score distillation to guide a text-to-3D optimization process to output visually plausible geometry and appearance.