Coordinator: Prof. Antonio Vicino
Home |  DIISM |   | Login Privacy e Cookie policy



Multimodal Signal Processing


Alessandro Mecocci
University of Siena - Dipartimento di Ingegneria dell'Informazione e Scienze Matematiche
Course Type
Type B
luglio 2008
Introduction: definition of Multi-Modal and Multi-Media systems, use of multiple heterogeneous sensors, complementarity, competitiveness, cooperativeness, independence. Physical sensors: sensor characterization, video, sound, temperature, flux, force, position, rotation, acceleration, speed. Abstract sensors: logical architecture, hierarchy, templates. Video processing: Low Level preprocessing (enhancement, filtering, restoration, segmentation), Mid-Level representation (ROIs, shape description, RAG, feature extraction and selection), Motion Analysis, Context Representation (boundary representations, Simplex grids, multidimensional trees), Context Awareness. Sound Processing: filtering, speech recognition, localization, acoustic target tracking, sound recognition. Multimodal Sensor Fusion: spatial calibration, temporal calibration, frequency calibration, arrays of audio and video sensors, Wireless Sensor Networks, Distributed Processing, JDL model, statistical approaches, Bayesian Networks, Dempster and Shafer theory, geometric approaches, Fuzzy theory, Rule-Based approaches. Application of Multimodal Systems: multisensory target recognition, multimodal motion detection and tracking, Activity Maps, actions interpretation, advanced video-surveillance applications (aggressive act detection, anomalous behavior detection, graffiti detection, abandoned items detection, Intelligent Transportation Systems), Multimodal Biometry, Assistive Environments for Elderly People and Individuals with Special Needs, SMART Rooms, Perceptual Environments, Social Interfaces for Cultural Heritage



PhD Students/Alumni

Dip. Ingegneria dell'Informazione e Scienze Matematiche - Via Roma, 56 53100 SIENA - Italy