Sensorimotor Control

Learning to Act by Predicting the Future

Project Overview

We are working on sensorimotor control in immersive environments.

Publications

Learning to Act by Predicting the Future

Learning Complex Neural Network Policies with Trajectory Optimization

Variational Policy Search via Trajectory Optimization

Guided Policy Search

Continuous Inverse Optimal Control with Locally Optimal Examples

Nonlinear Inverse Reinforcement Learning with Gaussian Processes

Feature Construction for Inverse Reinforcement Learning