The University of Southampton
Telephone:
+442380598866
Email:
gvv@it-innovation.soton.ac.uk

Dr Galina V Veres

Research Staff

Galina Veres is a Senior Research Engineer at IT Innovation Centre, ECS. She gained  a PhD from National University of Science and Technology in engineering in 1997 and is an expert in computer vision, human behaviour recognition, machine learning, data mining, artificial intelegence and automated state detection in complex systems. She joined Southampton University at 2002 as a Research Fellow, participating at first at the project concerend with developing global paralell optimisation techniques for distributed suction for boundary layer transition control, then participating in DTC- Biometeric Data Fusion for Secure Enviroments. She joined IT Innovation Centre in 2006 and took part in a number of EU funded projects covering different areas of research in the fields of pattern recognition and machien learning.

Research

Research interests

Video-based human behaviour recognition; machine learning techniques; automatic state detection both video-based and time-series based; artificial intelligence; statistical analysis;analysis of complex time series; prediction and estimation techniques.

Projects

Grants

RESEARCH PROJECTS

ZONESEC This project concerns with detection and classification of illicit behaviour using different sensors.  Robust Event Detector based on Median Absolute Deviation (MAD) was developed for detecting illicit behaviour using vibration sensors.  Additionally, the sensitivity of MAD Event detector to window size was investigated for high level event detection and event detection on window by window basis. Additionally, attempts were made to classify events between No-Activity, Kick and Shake using measurements from vibration sensors installed at fences for the protection of critical infrastructure.  Different domains (time, frequency, time-frequency and wavelet) and different classification algorithms (RF, DT, SVM, AdaBoost) were applied using vibration signal.  Distinguishing between Kick and Shake is challenging problem in this case due to low sampling rate, however we can distinguish between Activity, No-Activity, Start of Activity and End of Activity with good accuracy (above 80%). Further, the drones can help in detection illicit behaviour in WideZone. The work is done on detection of known structures, cars, people and illicit behaviours using videos provided by drones.

SIMMI-OBDA This project was an earlier research study concerned with automatic identification of classes during Electrofacies analysis for Hole 1256D using Decision Tree classification algorithms. Nine classes were identified in Volcanic Section by texture, presence/absence/distribution of fractures and ranges of conductivity.  Classification of classes is very challenging problem due to low recovery rate for some classes. Therefore classification were done in two stages: 3 classes with the highest recovery rate and all 9 classes. Though good results were obtained for 3 classes, 9 classes problem remain challenging. New features for each class were identified and extracted which allowed some improvement in classification results for 9 classes.  Overall, this initial study showed that it is possible to automatically classify volcanic rock, though some challenges are remaining with 9 class problem. Possible further research was identified such as updating features and using deep learning.

OPTET Bayesian inferences modelling for misbehaviour and threats in cyber security systems; formal definition of normal behaviour, natural variations and misbehaviours for cyber security systems and developing tests to detect misbehaviours due to protracted threat.

DAVID Developing risk analysis framework, risk modelling procedures and procedures dealing with risk for digital preservation. The risk modelling procedure allows to simulate both mutually exclusive risks and risks which can take place together. Procedures dealing with risks are based on the guidelines provided by ORF Austria.

DESURBS This project is dealing with designing safer urban spaces. Fear-meter data was analysed to identify safe areas. The results were compared with produced interpolation maps based on police crime data.

TRIDEC Operational state detection using machine learning techniques for drilling application, analysis of a complex time series data.  Different machine learning techniques were applied such as Random Forest, Support Vector Machines, kNN, LDA, ESN, AdaBoostM2, RUSBoost and Subspace. Additionally, recommendations were given how to select the most suitable approach for Operational States Detection.

SCOVIS This project dealt with automatic workflow monitoring in industrial environment using visual sensors.  The resulting technology allows easy installation of intelligent supervision systems. SCOVIS supports the automatic detection of behaviours, workflow violations, and localisation of salient moving or static objects in scenes monitored by multiple cameras.  The project investigates weakly supervised learning algorithms and self-adaptation strategies for analysis of visually observable workflows and behaviours. Camera network coordination is also supported so that complex behaviours can be identified as combination of spatio-temporal object relations in multiple scenes

PRESTOSPACE Correcting evaluation process of four pertinent defects occurring during tape preservation and storage; Using different classification scheme to extract the most problematic tapes given diffusion years and brands.  

SANY Developing an automatic spatial interpolation algorithm based on kriging which can be operated as a service and optimised for use in a maritime application and an application monitoring the geological effects of tunnelling activities.

Publications

Veres, G.V., Tutty, O.R., Rogers, E. and Nelson, P.A. (2008) Global-optimization-based methods for use in boundary layer transition control [in special issue: Active Flow Control] Proceedings of the I MECH E Part I Journal of Systems & Control Engineering, 222, (5), pp. 297-307. (doi:10.1243/09596518JSCE465).

Veres, G.V., Tutty, O.R., Rogers, E. and Nelson, P.A. (2003) Constrained parallel global optimization for boundary layer transition control In Proceedings of the 42nd IEEE Conference on Decision and Control. Institute of Electrical and Electronics Engineers., pp. 2792-2797. (doi:10.1109/CDC.2003.1273047).

Veres, G.V., Tutty, O.R., Rogers, E. and Nelson, P.A. (2004) Global optimisation-based control algorithms applied to boundary layer transition problems [in special issue: UKACC Conference Control 2002] Control Engineering Practice, 12, (4), pp. 475-490. (doi:10.1016/j.conengprac.2003.09.009).

Veres, G.V., Tutty, O.R., Rogers, E. and Nelson, P.A. (2003) On design optimisation based control methods for distributed suction for boundary layer transition control At Proceedings of the 2003 American Control Conference, United States. 04 - 06 Jun 2003. , pp. 2187-2192. (doi:10.1109/ACC.2003.1243340).

Veres, G.V., Tutty, O.R., Rogers, E. and Nelson, P.A. (2003) Parallel global optimisation based control of boundary layer transition At European Control Conference (ECC'03), United Kingdom. 01 - 04 Sep 2003. 6 pp.

Veres, Galina, Gordon, Layla, Carter, John and Nixon, Mark (2004) What image information is important in silhouette-based gait recognition At IEEE Computer Vision and Pattern Recognition conference.

Veres, Galina V, Nixon, Mark S, Middleton, Lee and Carter, John N. (2005) Fusion of dynamic and static features for gait recognition over time At Eighth International Conference of Information Fusion, United States. 25 - 29 Jul 2005.

Veres, Galina, Nixon, Mark and Carter, John, (2005) Modelling the time-variant covariates for gait recognition Kanade, T, Jain, A.K. and Ratha, N.K. (eds.) At AVBPA2005, Lecture Notes in Computer Science,, United States. 20 - 22 Jul 2005. , pp. 597-606.

Veres, Galina, Nixon, Mark and Carter, John (2005) Model-based approaches for predicting gait changes over time At International Workshop on Pattern Recognition.

Veres, Galina V., Huynh, Trung Dong, Nixon, Mark S., Smart, Paul R. and Shadbolt, Nigel R. (2006) The Military Knowledge Information Fusion Via Semantic Web Technologies s.n.

Zlatev, Z., Middleton, S.E. and Veres, G. (2009) Ordinary kriging for on-demand average wind interpolation of in-situ wind sensor data At EWEC 2009, France. 16 - 19 Mar 2009. 7 pp.

Veres, Galina, Grabner, Helmut, Middleton, Lee and Van Gool, Luc (2010) Automatic Workflow Monitoring in Industrial Environments At Asian Conference on computer Vision (ACCV), New Zealand. 10 - 12 Nov 2010.

Zlatev, Zlatko, Middleton, Stuart and Veres, Galina (2010) Benchmarking Knowledge-assisted Kriging for Automated Spatial Interpolation of Wind Measurements At Fusion 2010.

Veres, Galina, Middleton, Lee, Sabeur, Zoheir, Bouchrika, Imed, Arbab-Zavar, Banafshe, Carter, John, Nixon, Mark, Grabner, Helmut, Stalder, Severin, Van Gool, Luc, Morzinger, Roland, Thaler, Marcus, Rene, Schuster, Albert, Hofman and Thallinger, Georg (2010) Tools for semi-automatic monitoring of industrial workflows At ACM Multimedia 2010, Italy. 25 - 29 Oct 2010.

Zlatev, Zlatko, Middleton, Stuart, Veres, Galina, Salvo, Nicola and Sabeur, Zoheir (2010) Generic Sensor Data Fusion Services forWeb-enabled Environmental Risk Management and Decision-Support Systems At EGU General Assembly.

Voulodimos, Athanasios, Kosmopoulos, Dimitrios, Veres, Galina, Grabner, Helmut, Van Gool, Luc and Varvarigou, Theodora (2011) Online classification of visual tasks for industrial workflow monitoring Neural Networks

Veres, Galina and Sabeur, Zoheir (2013) Automated operational states detection for drilling systems control in critical conditions At ESANN 2013, Belgium. 24 - 26 Apr 2013. 6 pp.

Zlatev, Zlatko, Veres, Galina and Sabeur, Zoheir (2013) Agile data fusion and knowledge base architecture for critical decision support International Journal of Decision Support System Technology (IJDSST), 5, (2), pp. 1-20.

Bailer, Werner, Hall-May, Martin and Veres, Galina (2014) Metadata representation and risk management framework for preservation processes in AV archives At 11th International Conference on Digital Preservation (iPres), Australia. 06 - 10 Oct 2014. 3 pp.

Engen, Vegard, Veres, Galina, Hall-May, M., Chenot, Jean-Hugues, Bauer, Christoph, Bailer, Werner, Hoffernig, Martin and Houpert, Jorg (2015) DAVID D3.3 - Final IT-based tools and strategies for avoiding, mitigating and recovering from digital AV loss & final conceptual risk management framework and tools for digital AV preservation , Southampton, GB University of Southampton 92pp.

Veres, Galina and Sabeur, Zoheir (2015) Data analytics for drilling operational states classifications At European Symposium on Artificial Networks, Computational Intelligence and Machine Learning (ESANN2015), Belgium.

Engen, Vegard, Veres, Galina, Crowle, Simon, Bashevoy, Maxim, Walland, Paul and Hall-May, Martin (2015) A Semantic Risk Management Framework for Digital Audio-Visual Media Preservation At The Tenth International Conference on Internet and Web Applications and Services (ICIW), June 21 - 26, 2015, Brussels, Belgium, Belgium.

Engen, Vegard, Veres, Galina, Crowle, Simon, Walland, Paul and Bauer, Christoph (2016) Business process risk management and simulation modelling for digital audio-visual media preservation International Journal On Advances in Internet Technology, 9, (1 & 2), pp. 12-30.

Sabeur, Zoheir, Zlatev, Zlatko, Melas, Panagiotis, Veres, Galina, Arbab-Zavar, Banafshe, Middleton, Lee and Museux, Nicolas (2017) Large scale surveillance, detection and alerts information management system for critical infrastructure At International Symposium on Environmental Software Systems, Zadar, Croatia. 10 - 12 May 2017. 10 pp.

Sabeur, Z.A., Correndo, G., Veres, G., Arbab-Zavar, B., Lorenzo, J., Habib, T., Haugommard, A., Martin, F., Zigna, J.-M. and Weller, G. (2017) EO Big Data connectors and analytics for understanding the effects of climate change on migratory trends of marine wildlife At International Symposium on Environmental Software Systems, Zadar, Croatia. 10 - 12 May 2017. 9 pp.

Contact

Share this profile FacebookGoogle+TwitterWeibo

We use cookies to ensure that we give you the best experience on our website. If you continue without changing your settings, we will assume that you are happy to receive cookies on the University of Southampton website.

×