An Introduction to Action Recognition and Tracking in Videos

in
PhD Seminar Course on

An Introduction to Action Recognition and Tracking in Videos

Cagliari, April, 12-20, 2011
Instructor: Prof. Massimo Piccardi
Massimo Piccardi, M.Eng. in EE (University of Bologna, Italy), Ph.D. in CE/CS (University of Bologna, Italy).

Massimo Piccardi is a Professor of Computer Systems with the Faculty of Engineering and Information Technology at University of Technology, Sydney (UTS) that he joined in January 2002 as an associate professor. Previously, he was a senior lecturer/assistant professor with the Faculty of Engineering at the University of Ferrara, Italy. At UTS, he serves a Director of the Advanced Video Surveillance research program. His main research interests are in the areas of pattern recognition, computer vision, image and video analysis and processing, and computer architecture, with main applications to video surveillance, multimedia, and human-computer interaction (CHI). He has been the author or co-author of more than 120 scientific papers on international journals and conferences, and three book chapters. He has been the leading investigator in several research projects including two ARC (Australian Research Council) Discovery Projects and an ARC Linkage Project. Dr. Piccardi is a senior member of the IEEE and the IEEE Computer Society and a member of the International Association for Pattern Recognition. He is a member of the Steering Committee of the IEEE International Conference series on Advanced Video and Signal-based Surveillance (AVSS) and serves as an Associate Editor of journals Image and Vision Computing and Machine Vision and Applications.
Duration: 13 hours
Schedule:
  • Tuesday 12 April 2011 11:00-12:30; 14-15:30
    • Recalls of probability theory.
    • Main properties of the Gaussian distribution.
    • Gaussian mixture models.
  • Thursday 14 April 2011 11:00-12:30; 14-16:30
    • Concepts of Bayesian classification.
    • Density estimation.
    • Parametric and non-parametric estimators.
  • Monday 18 April 2011 11:00-12:30; 14-15:30
    • The hidden Markov model.
    • Action recognition.
  • Wednesday 20 April 2011 11:00-12:30; 14-16:30
    • Tracking in videos.
    • Recursive Bayesian estimation.
    • The Kalman filter.
    • Particle filters.
Venue:
Aula piano rialzato – DIEE - Padiglione A
Topics:

Objectives

This course will introduce the students to techniques for action recognition and tracking in videos. The first few lectures offer recalls of foundational topics such as probability, Bayesian classification and density estimation which will facilitate the comprehension of the following topics. Action classification covers the hidden Markov model, bag-of features approaches and conditional random fields. Tracking covers recursive Bayesian estimation and the Kalman and particle filters. The course aims to be of benefit for research students in the areas of computer vision, pattern recognition, multimedia, image and signal processing, human-computer interaction, and also computer science/computer engineering in general.

 

Learning modalities

The course will be taught over 4 days in eight slots, two slots per day with a break in between. The presentation will be based on slides and occasional use of whiteboard. The course can be taught in either Italian or English at the preference of the convenor.


Materials
A copy of the slides can be downloaded from here. The copyright of all materials stays with the copyright holders.

Topics at a glance

  • Recalls of probability theory
  • Concepts of Bayesian classification
  • Density estimation
  • The hidden Markov model 
  • Action recognition
  • Tracking in videos.
  • Recursive Bayesian estimation.
  • The Kalman filter.
  • Particle filters
Organizer: Prof. Fabio Roli
Dep. of Electrical and Electronic Engineering
University of Cagliari, Italy
Email: roli(at)diee(dot)unica(dot)it