• Start Time:
  • End Time:
  • Day:
    Day 1


Disney Research has developed an automatic “ghosting” system which illustrates where defensive players should have been (instead of where they actually were) based on the locations of the opposition players and ball. Employing a machine learning technique called deep imitation learning, and modifying standard recurrent neural network training to consider both instantaneous and future losses, their algorithm is able to generate ghost player behaviors which anticipate the movements of their teammates and the opposition. The approach avoids the man-years of manual annotation that is needed to train existing ghosting systems, and can be fine tuned to mimic the behavior of specific teams or playing styles.

Associated Speakers:

Peter Carr

Research Scientist

Disney Research

Associated Talks:

11:30AM - Day 1

View Case Study: Modeling Sequential Decision Making in Team Sports using Deep Imitation Learning

View Full Info

Powered by WishList Member - Membership Software