Unsupervised Learning of Object Structure and
          Dynamics from Videos

        Matthias Minderer Chen Sun Ruben Villegas Forrester Cole
                Kevin Murphy Honglak Lee
                   Google Research
      {mjlm, chensun, rubville, fcole, kpmurphy, honglak}@google.com

                        Abstract
      Extracting and predicting object structure and dynamics from videos without
      supervision is a major challenge in machine learning. To address this challenge,
      we ...