Vous êtes ici

Learning to Move: Recent Progress Towards Physics-based Characters

Jeudi, 1. décembre 2016 - 11:00 - 11:45
Michiel van de Panne, University of British Columbia
IRISA Rennes, Salle Michel Métivier

Dans le cadre de la soutenance de thèse d'Ana Lucia Cruz Ruiz le 2 décembre, Charles Pontonnier et Georges Dumont de l'équipe MimeTIC organisent  le 1er décembre à 11h en salle Michel Métivier, un séminaire de Michiel Van de Panne, professeur à University of British Columbia, Vancouver. Il est spécialiste de l'animation de personnages basée physique.


Learning to Move:  Recent Progress Towards Physics-based Characters 

Michiel van de Panne 
University of British Columbia 

Interactive physics-based characters are now capable of reproducing a growing number of motion skills, often with a focus on generating agile-and-robust locomotion.  In this talk, I review recent progress in simulation-based models of human and animal motion as used for computer animation, where they seek to replace simpler kinematic models based on motion-capture.  I'll discuss the roles of simplified dynamical models, optimization, and machine learning (deep learning and not-so-deep learning) in these approaches. as well as what insights our simulation-based work in animation may provide for biomechanics and robotics.  A wide variety of results will be shown to illustrate the capabilities of current methods.  I'll also identify several research directions where still require significant work.