20 papers with code • 0 benchmarks • 0 datasets
These leaderboards are used to track progress in Normalising Flows
Data-driven modelling and synthesis of motion is an active research area with applications that include animation, games, and social robotics.
We show that normalising flows become pathological when used to model targets whose supports have complicated topologies.
Variational inference has had great success in scaling approximate Bayesian inference to big data by exploiting mini-batch training.
Generative flows are promising tractable models for density modeling that define probabilistic distributions with invertible transformations.
In this paper, we introduce Woodbury transformations, which achieve efficient invertibility via the Woodbury matrix identity and efficient determinant calculation via Sylvester's determinant identity.
In interactive scenarios, systems for generating natural animations on the fly are key to achieving believable and relatable characters.