Home      Log In      Contacts      FAQs      INSTICC Portal
 

Keynote Lecture

 

Enhanced Protection of Vulnerable Road Users – A combined discriminative and generative approach for accurate detection and prediction of pedestrian intentions

Miguel A. Sotelo
Universidad de Alcalá
Spain
 

Brief Bio
Miguel Ángel Sotelo received the degree in Electrical Engineering in 1996 from the Technical University of Madrid, the Ph.D. degree in Electrical Engineering in 2001 from the University of Alcalá (Alcalá de Henares, Madrid), Spain, and the Master in Business Administration (MBA) from the European Business School in 2008. From 1993 to 1994, he held an Excellence Research Grant at the University of Alcalá, where he is currently a Full Professor at the Department of Computer Engineering and Vice-president for International Relations. In 1997, he was a Research Visitor at the RSISE of the Australian National University in Canberra. His research interests include Self-driving cars, Cooperative Systems, and Traffic Technologies. He is author of more than 200 publications in journals, conferences, and book chapters. He has been recipient of the Best Research Award in the domain of Automotive and Vehicle Applications in Spain in 2002 and 2009, and the 3M Foundation Awards in the category of eSafety in 2004 and 2009. He served as Auditor and Expert at FITSA Foundation for RTD Projects in the domain of automotive applications in 2004-2010. Miguel Ángel Sotelo has served as Project Evaluator, Rapporteur, and Reviewer for the European Commission in the field of ICT for Intelligent Vehicles and Cooperative Systems in FP6 and FP7. He was Director General of Guadalab Science & Technology Park (2011-2012) and co-founder and CEO of Vision Safety Technologies (2009-2015), a spin-off company established in 2009 to commercialize computer vision systems for road infrastructure inspection. He is member of the IEEE ITSS Board of Governors and Executive Committee. Miguel Ángel Sotelo served as Editor-in-Chief of the Intelligent Transportation Systems Society Newsletter (2013), Editor-in-Chief of the IEEE Intelligent Transportation Systems Magazine (2014-2016), Associate Editor of IEEE Transactions on Intelligent Transportation Systems (2008-2014), member of the Steering Committee of the IEEE Transactions on Intelligent Vehicles (since 2015), and a member of the Editorial Board of The Open Transportation Journal (2006-2015). He has served as General Chair of the 2012 IEEE Intelligent Vehicles Symposium (IV’2012) that was held in Alcalá de Henares (Spain) in June 2012. He was recipient of the 2010 Outstanding Editorial Service Award for the IEEE Transactions on Intelligent Transportation Systems, the IEEE ITSS Outstanding Application Award in 2013, and the Prize to the Best Team with Full Automation in GCDC 2016. At present, he is President of the IEEE Intelligent Transportation Systems Society.  


Abstract
Driver Assistance Systems have achieved a high level of maturity in the latest years. As an example of that, sophisticated pedestrian protection systems are already available in a number of commercial vehicles from several OEMs. However, accurate pedestrian path prediction is needed in order to go a step further in terms of safety and reliability, since it can make the difference between effective and non-effective intervention. Getting to understand the underlying intent of an observed pedestrian is of paramount interest in a large variety of domains that involve some sort of collaborative and competitive scenarios, such as robotics, surveillance, human-machine interaction, and intelligent vehicles. In contrast to trajectory-based approaches, the consideration of the whole pedestrian body language has the potential to provide early indicators of the pedestrian intentions, much more powerful than those provided by the physical parameters of a trajectory. In this talk, we consider a hybrid approach in which Deep Learning techniques are used to robustly detect pedestrians’ body parts and pose together with a generative approach, based on GPDM (Gaussian Process with Dynamical Model), for accurate trajectory prediction in a time horizon of up to 1.0 s. The proposed system constitutes a further step in the state-of-the-art in the quest for advanced VRU protection systems. 



footer