Temporal Learning in Video Data Using Deep Learning and Gaussian Processes

This paper presents an approach for data-driven modeling of hidden, stationary temporal dynamics in sequential images or vidoes using deep learning and Bayesian non-parametric techniques. In partic