SIGGRAPH 2017 Learning to Move

DeepLoco: Dynamic Locomotion Skills Using Hierarchical Deep Reinforcement Learning

With hierarchical deep reinforcement learning, this work develops a variety of environment-aware locomotion skills for a simulated 3D biped.

Xue Bin Peng
The University of British Columbia

Glen Berseth
The University of British Columbia

KangKang Yin
National University of Singapore

Michiel van de Panne
The University of British Columbia

Phase-Functioned Neural Networks for Character Control

This paper proposes a real-time character control mechanism using a novel phase-functioned neural network that can learn how humans move through complex terrains while avoiding obstacles.

Daniel Holden
University of Edinburgh

Taku Komura
University of Edinburgh

Jun Saito
Method Studios

Learning to Schedule Control Fragments for Physics-Based Characters Using Deep Q-Learning

A deep Q-learning-based method for learning a scheduling scheme that reorders short control fragments as necessary at runtime to achieve robust control of challenging skills such as skateboarding.

Libin Liu
Disney Research Pittsburgh

Jessica Hodgins
Carnegie Mellon University

Discovering and Synthesizing Humanoid Climbing Movements

With this path planner and movement optimizer for a simulated humanoid climber, the state of the art is advanced through enabling rich, dynamic movements where more than one limb can move at the same time, and limbs can also hang free for balance, or use the wall around the holds for friction.

Kourosh Naderi
Aalto University

Joose Rajamäki
Aalto University

Perttu Hämäläinen
Aalto University

Search form

You are here

Learning to Move

DeepLoco: Dynamic Locomotion Skills Using Hierarchical Deep Reinforcement Learning

Phase-Functioned Neural Networks for Character Control

Learning to Schedule Control Fragments for Physics-Based Characters Using Deep Q-Learning

Discovering and Synthesizing Humanoid Climbing Movements

Exhibit at SIGGRAPH 2017

Sponsorship Opportunities

Registration Details