Guilherme de Souza


Guilherme de Souza
Keynote Speaker - University of Missouri, USA

Talk: Are we approaching a Third Winter in AI? A brief history on the mistakes made and how to avoid making them again.

I am an Associate Professor in Electrical Engineering and Computer Science, an Adjunct Professor with the Sinclair School of Nursing, and a core faculty member of the MU Informatics Institute at the University of Missouri (MU). I am also the Director of the Vision-Guided and Intelligent Robotics Laboratory (ViGIR Lab), which currently consists of six PhD, two MS and two undergraduate students. I published over 90 refereed articles in 3D modeling, robotic vision, mobile robot navigation, pattern recognition and computational intelligence. My papers are all peer-reviewed and were published in many journal venues like the IEEE Transactions on Pattern Analysis and Machine Intelligence and the Journal of Computer Vision and Image Understanding, as well as in high-impact conferences and workshops with acceptance rates as low as 28%, such as BMVC, ECCV, ICRA, etc. My current and past research projects have been mainly funded by the NSF (including a $20M grant to a consortium of universities and research centers), NIH (R15 and R01), DoD, the National Geospatial-Intelligence Agency, the Naval Research Laboratory, the US Department of Agriculture, the Missouri Soybean Merchandising Council, etc. While varied in application, my research is mainly focused on building, recognizing and applying data models to represent the world as perceived by different types of robots through their sensors. I am particularly interested in the extraction and recognition in real time of patterns – visual or otherwise (e.g. 2D and 3D imaging, sEMG, etc). My research has helped improve early detection of lymphedema using 3D modeling running on smatphones; dysphagia (swallowing disorders) and voice disorders using pattern recognition of sEMG signals obtained in the lab or from portable/wearable devices; and plant phenotyping and genotyping using field robotics.  Research in my lab, the Vision Guided and Intelligent Robotics Lab (ViGIR), has also allowed the development of technology to help people with severe disability to drive wheelchairs; to phenotype shoots and roots of plants growing in hydroponic substrate; and to create models of the human body – e.g. tracking and detection of human activity for military survailance. (Webpage)