Rezha Adrian Tanuharja

Research Focus

Generative Artificial Intelligence on Non-Euclidean Manifolds

Description

Generative AI has gained significant attention recently, with advancements enabling synthetic data generation across various domains. However, these models often assume data lies in Euclidean spaces with flat geometry, while real-world data frequently resides on low-dimensional Riemannian manifolds within higher-dimensional spaces. This research aims to investigate models that account for these underlying structures to improve performance and efficiency.