Vincenzo Caglioti obtained a Laurea degree (magna cum laude) in Electronics Engineering from Politecnico di Milano in 1987. He obtained a Phd in Computer and System Engineering in 1992 from Politecnico di Milano. In 1993, he visited International Computer Science Institute at Berkeley (CA, USA) as a post-doc. From 1995 to 2001 he was Assistant Professor (“ricercatore”) at the Politecnico di Milano. From 2002 he is an Associate Professor at the Politecnico di Milano. Teaching activity in University courses: Computer Programming, Foundations of Computer Engineering, Information Systems, Signal Processing, Knowledge Engineering and Expert Systems, Robotics, Image analysis and Synthesis, Computer Vision, Advanced Topics in Image Analysis. Teaching activity in International Master Courses, or International Exchange Courses: “Soft Computing in Vision” (International Master on Soft Computing for Industrial Applications), “Multiview Geometry and Applications to Computer Vision” (Athens Course). Invited talks: UC Berkeley (CA, USA), McGill University (Canada), University of Verona (Italy).
Main research interests are in Image Analysis and Computer Vision, and in Robotics. Generally, addressed problems consist in the automatic construction of models of an observed scene, or model of a robot environment, through both passive techniques (analysis of the data acquired through visual sensors or other robot sensors) and active techniques: the latter ones involve the planning of sensor exploration strategies in order to acquire as much as information is possible. Sensor exploration strategies for localization and recongnition tasks have been extensively studied, while taking into account the compatational complexity involved in planning exploration strategies. A system for cooperative localization in multi-robot systems has been constructed, based on optimal exploration strategies. Methodologies for visual object recognition have been developed, while proving theoretical results concerning computational complexity of recognition. A robot mapping system has been realized, based on the optimization of both the exploration path and the sensor operations. Active exploration has also been applied to environmental monitoring (e.g., localization of magnetic field sources). 3D shape reconstruction from single and multiple images has been studied. In particular, gemetrical properties related to noncentral catadioptric cameras have been proved, and reconstruction methods of straight lines in indoor scenes from single catadioptric images have been developed. Reconstruction methods for deformable objects (e.g. paper-like surfaces) have been realized. Techniques for visual analysis of sport events or other complex dynamic scenes have also been developed (e.g. to determine non-parabolic trajectories of soccer ball, or to reconstruct the ball spin. The relationships between motion blur and transparency have been studied, and techniques have been derived for determining image motion and temporal super-resolution from blurred images (e.g., estimating parameters relative to string vibrations in music instruments). Sub-pixel image analysis techniques have been developed, and applied to the accurate and fast reconstruction of moving panels in automated sanding machines by a simple device base on a camera plus a shadowng screen. Human gesture recongnition is studied, in order to develop an embedded camera systems for detecting crime situations.
Research results have been published in 17 papers in international journals, 34 papers in international conference proceedings or international books. Participation in International Funded Research Projects: Automatic Error Recovery in Programmed Robots (IBM), Multisensor system for Robotics (EU – Esprit), Robot Integration into CIM (EU – Esprit), Crowd Monitoring (EU - IT), Optical Measuring Systems for Panel Pose and Shape (EU – VI FP) Other national funded projects: 3D vision in robot guidance (Italian Ministry for Scientific Research), Robot Programming and Planning Systems (Italian National Research Institute (CNR)), Visual object localization and Recognition (CNR), Robotic unit for Assistance to Disabled Persons (CNR), Parallel Processing for Robot Control (Italian Institute for Alternative Energy). Visual Measurement of Velocity Field in Aluminium Extrusion (Compes), Visual Analysis of reflecting Surfaces (Bavelloni), Embedded Camera System for the Detection of Crime Events (Selea). Participation to Program Committees: 2006 Robocup Symposium, 2007 Workshop on Robot Vision, 2007 Conference on Visual Information Enginnering, 2007 Robocup Symposium, 2007 IASTED Conference on Visualization, Imaging, and Image processing, 2008 Workshop on Robot Vision. Reviewer for International Journals: IEEE Transactions on Robotics, IEEE Transactions on Pattern Analysis and Machine Intelligence, IEEE Transactions on Systems, Man, and Cybernetics, IEEE Transactions on Instrumentation and Measurement, IEEE Transactions on Neural Networks, Pattern Recognition, Pattern Recognition Letters, International Journal of Computer-Integrated Manufacturing, International Journal of Robotics and Autonomous Systems, International Journal of Robotics and Automation, International Journal of Intelligent and Fuzzy Systems, International Journal of Intelligent and Robotic Systems. International Project Review activity: reviewer for Belgian Ministry of Scientific Research for Projects in the filed of Robotics.