Deep learning models related to space and time (trajectories, video, etc).

Content:

1. Gluon implementation of action in video recognition article "Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset" by Deepmind (translated from TensorFlow).
2. Gluon implementation of trajectory prediction in "Lorenz Map Prediction: A Trip Through Time" . (from scratch)