Normalising Flows

Normalising Flows are a type of probabilistic model that maps complex distributions to simpler ones through a series of invertible transformations, enabling efficient sampling and density estimation. The goal is to construct flexible and interpretable generative models that can capture the intricate structures within data. Normalising Flows have significant applications in high-dimensional data modeling, image generation, speech synthesis, and other fields, enhancing the expressiveness and quality of generated outputs.