A Survey on Reinforcement Learning Methods in Character Animation

Eurographics 2022

Ariel Kwiatkowski¹, Eduardo Alvarado¹, Vicky Kalogeiton¹, Karen Liu², Julien Pettré³, Michiel van de Panne⁴, Marie-Paule Cani¹
¹LIX, Ecole Polytechnique/CNRS, Institut Polytechnique de Paris, Palaiseau, France
²Stanford University, Stanford, CA, USA
³Univ Rennes, Inria, CNRS, IRISA, Rennes, France
⁴University of British Columbia, Vancouver, Canada


Reinforcement Learning is an area of Machine Learning focused on how agents can be trained to make sequential decisions, and achieve a particular goal within an arbitrary environment. While learning, they repeatedly take actions based on their observation of the environment, and receive appropriate rewards which define the objective.

This experience is then used to progressively improve the policy controlling the agent’s behavior, typically represented by a neural network. This trained module can then be reused for similar problems, which makes this approach promising for the animation of autonomous, yet reactive characters in simulators, video games or virtual reality environments. This paper surveys the modern Deep Reinforcement Learning methods and discusses their possible applications in Character Animation, from skeletal control of a single, physically-based character to navigation controllers for individual agents and virtual crowds. It also describes the practical side of training DRL systems, comparing the different frameworks available to build such agents.


    	title = {A Survey on Reinforcement Learning Methods in Character Animation},
    	author = {Kwiatkowski, Ariel and Alvarado, Eduardo and Kalogeiton, Vicky and Liu, C. Karen and Pettr{\'{e}}, Julien and van de Panne, Michiel and Cani, Marie-Paule},
    	doi = {10.48550/arxiv.2203.04735},
    	eprint = {2203.04735},
    	journal = {Computer Graphics Forum},
    	publisher = {{Wiley}},
    	volume = {41},
    	pages= {1-27},
    	year = {2022},
    	month = {mar},
    	url = {https://hal.inria.fr/hal-03600947},