Probabilistic Parametric Curves for Sequence Modeling

• Main Author: Hug, Ronny (auth)
• Format: Electronic Book Chapter
• Language: English
• Published: Karlsruhe KIT Scientific Publishing 2022
• Series: Karlsruher Schriften zur Anthropomatik
• Subjects: Maths for computer scientists; Probabilistische Sequenzmodellierung; Stochastische Prozesse; Neuronale Netzwerke; Parametrische Kurven; Probabilistic Sequence Modeling; Stochastic Processes; Neural Networks; Parametric Curves
• Online Access: DOAB: download the publication; DOAB: description of the publication

This work proposes a probabilistic extension to Bézier curves as a basis for effectively modeling stochastic processes with a bounded index set. The proposed stochastic process model is based on Mixture Density Networks and Bézier curves with Gaussian random variables as control points. A key advantage of this model is given by the ability to generate multi-mode predictions in a single inference step, thus avoiding the need for Monte Carlo simulation.