Virginia Fernandez Arguedas

Contact Details

Research Group: Multimedia and Vision Research Group

Supervisor: Prof Ebroul Izquierdo

Research Topic: Object Classification for Surveillance Videos

Research Scope: The recent exponential popularity of surveillance systems could be largely attributed to increased safety concerns of the public. The CCTV ubiquitousness generates everyday a large amount of video data which requires constant human supervision. In efforts to mitigate the dependency of constant supervision of surveillance footage, several object and event detection and recognition algorithms has been developed in recent times. Generally the proposed algorithms focus their analysis on the visual appearance of the objects. However, psychological studies have shown that human beings can routinely discriminate and recognise the type of object using motion or behaviour patterns, even in large viewing distances where the scene of observation is affected by either poor visibility conditions or in circumstances where other familiarity cues such as appearance is hard to distinguish. Hence, this research concentrates on developing a framework for robust and efficient object classification for surveillance videos based not only in the visual appearance of the detected moving objects but also on their general behaviour.

Research interests: Object classification, behaviour analysis, visual analysis, pattern recognition, object tracking, machine learning, semantic analysis and retrieval of multimedia content.

Participation in projects

Biography

I was born in Madrid, Spain, in 1984. I received my BSc and MSc in Telecommunications Engineering from Universidad Autonoma de Madrid in 2008. From September 2007 until June 2008, I developed my final Master Thesis at Queen Mary, University of London, in the Erasmus Mundus Program. Currently, I am a PhD student at Queen Mary, University of London, at Multimedia and Vision Research Group, working on Object Classification for Surveillance Videos. My research interests include pattern recognition, object classification, behaviour analysis, machine learning, object tracking, semantic analysis and retrieval of multimedia content. During my PhD, I have actively collaborated in some European funded projects, including PAPYRUS and VideoSense.

Publications:

Conferences

• V. Fernandez Arguedas, K. Chandramouli and E. Izquierdo, "Study of Particle Swarm Optimisation as Surveillance Object Classifier," in 3rd Latin American Conference on Networked and Electronic Media, 2011.
• V. Fernandez Arguedas and E. Izquierdo, "Object Classification based on Behaviour Patterns," in 4th Imaging for Crime Detection and Prevention, 2011.
• V. Fernandez Arguedas, K. Chandramouli, Q. Zhang and E. Izquierdo, "Optimal Combination of Low-Level Features for Surveillance Object Retrieval," in International Joint Conference on e-Business and Telecommunications, 2011.
• V. Fernandez Arguedas, K. Chandramouli, and E. Izquierdo, "Behaviour-based Object Classifier for Surveillance Videos," in 1st International Workshop on Eternal Systems, 2011.
• V. Fernandez Arguedas, Q. Zhang, K. Chandramouli and E. Izquierdo, "Multi-Feature Fusion for Surveillance Video Indexing," in 12th International Workshop on Image Analysis for Multimedia Interactive Services, 2011.
• V. Fernandez Arguedas, K. Chandramouli, and E. Izquierdo, "Semantic Object based Retrieval from Surveillance Videos," in 4th IEEE International Workshop on Semantic Media Adaptation and Personalization, 2009.
• V. Fernandez Arguedas, K. Chandramouli, and E. Izquierdo, "Exploiting Complementary Resources for Cross-Discipline Multimedia Indexing and Retrieval," in 1st International ICST Conference on User Centric Media, 2009.