Marco Cannici is a PhD student at Politecnico di Milano, where he received both the bachelor’s and master’s degrees in Computer Science and Engineering. Starting from his master thesis he worked on computer vision and event-based cameras. His research involved the development of new approaches for processing events that leverage conventional neural models. In particular he is interested in the design of frame-free architectures able to intrinsically exploit the sparse nature of the event-based visual encoding, mechanisms that use events to supervise conventional neural network in focusing on the salient portion of the scene and effective event representations able to adapt to the scene characteristics. He is now interested in an emerging research field of robotics, namely structured learning and code auto differentiation. The research in this field focuses in filling the gap between model-based approaches, leveraging prior knowledge in the form of robust algorithms and well understood system decompositions, and model-free approaches, relying instead on end-to-end training with powerful function approximators. Structured learning aims to leverage the benefits of both approaches, combining data-driven methods, mostly deep neural networks, with robust differentiable algorithms for filtering, planning and control.