Visual object tracking under fast motions can be hindered by conventional cameras, which are prone to producing blurry images. Event cameras are better suited for this task, but applying them to object tracking is non-trivial. We propose a framework to take advantage of these sensors by using deep learning techniques.