Image and Signal Analysis Group

Print
VSA grupės darbuotojai

 

 

Address
Akademijos st. 4, room 623
Vilnius, Lithuania

Head of the Group
dr. Povilas Treigys

Tel: +370 5 210 9328
E-mail:  

Lab on social networks:

Image and Signal Analysis Group

 

 

Research fields

 Speech and language processing:

  • Speech signal modelling: autoregressive/linear prediction, fractal-based nonlinear modelling, source-filter models.
  • Speech recognition: segment-based speech recognition.
  • Speech emotion recognition: hierarchical classification of speech emotions, speech emotion features, feature selection.
  • Estimation of speech quality: acoustic analysis for quality estimation; estimation of vocal fold functionality; estimation of phonation level;
  • Signal model parameter assessment;
  • Development of sound mathematical models.

 Image and video signal processing:

  • Object recognition and segmentation using deep neural networks
  • Image reconstruction by deep generative networks
  • Analysis of ocular fundus optic disc, other lesions
  • Analysis of images of colonic pathology to describe tumor cell subpopulations in whole-slide images
  • Detection of prostate changes in multiparametric magnetic resonance imaging
  • Processing of full-frame histological images by machine learning methods.

Analysis of big data and development of deep learning algorithms:

  • Analysis and recognition of clonal structures in genetic data.
  • Analysis of traffic (marine) data to identify traffic anomalies in LSTM and transformer deep neural networks
  • Development of deep teaching/learning methods for solving classification and forecasting problems.

 

Science and Education

Phd Students

Name, Surname Supervisor Field of
study		 Theme Time of study
Bernardas Čiapas Assoc. Prof. Dr. Povilas Treigys N 009 Image Feature Analysis for Solving Recognition Tasks in Self Checkout Systems 2019-2023
Monika Danilovaitė Assoc. Prof. Dr. Gintautas Tamulevičius N 009 Voice Quality Assessment Methodology for the Detection of Vocal Folds Disorders 2020-2025
Robertas Jurkus Assoc. Prof. Dr. Povilas Treigys T 007 Maritime Traffic Awareness Evaluation using Deep Neural Networks 2020-2024
Aleksas Vaitulevičius Assoc. Prof. Dr. Povilas Treigys N 009 Research and Modeling of Digital Signals for the Detection of Cancerous Areas of Tissue in Different MRI Modalities 2021-2025
Sandra Virbukaitė Dr. Jolita Bernatavičienė T 007 Development of Deep Learning Methods for Pathological Changes Identification in Eye Fundus Images 2020-2024
Roman Surkant Dr. Jolita Bernatavičienė N 009 Development of machine learning and functional data analysis methods for prostate multiparametric MRI analysis 2022-2026

 

Defended PhD thesis:

 

Defended MSc thesis:

  • "Speech Analysis-based Assessment of Vocal Folds State", 2020. Supervisor: dr. Gintautas Tamulevičius.
  • "Neural network based segmentation and classification of cell nuclei in whole slide histology images", 2020. Supervisor: dr. Povilas Treigys
  • "Fractal Modeling of Speech Signals", 2021. Supervisor: dr. Gintautas Tamulevičius.
  • "Outlier Detection in Multidimensional Streaming Data", 2021. Supervisor: dr. Jolita Bernatavičienė.
  • "Autoencoder-based Modelling of Speech Signals", 2021. Supervisor: dr. Gintautas Tamulevičius.
  • Šnekos signalų modeliavimas autoenkoderio tipo neuroniniais tinklais, 2021. Supervisor: dr. Gintautas Tamulevičius.
  • Investigation of Eye Fundus Image Quality on Vascular Segmentation Using Deep Neural Networks. 2022. Supervisor: dr. Jolita Bernatavičienė.
  • Evaluation Of Dynamic Contrast in Prostate MRI for Cancerous Tissue Identification. 2022.Supervisor:dr. Jolita Bernatavičienė.
  • Forecast of Energy Produced by Solar Power Plants. 2022. Supervisor: dr. Jolita Bernatavičienė.

 

Projects

  • Lithuanian state-funded project "Information Technologies for human Wellbeing - Clinical Decision support (E-Health)" 2003-2009.
  • Lithuanian state-funded project "Support of Clinical Decision Systems for e. Health System" 2007-2009.
  • Lithuanian state Science Foundation project "Development of Specialized data Analysis Methods for Heart Thermal Anisotropy Investigation" 2008.
  • Research service "Creation of a Model for the Recognition of Individual Lithuanian Words and Phrases" 2012-2013.
  • Researchers group project "Development and Inspection of the Prototype for the People with Disabilities Managed by Lithuanian Speech" 2012-2014.
  • EU  Structural funds project "Lithuanian Speech Guided Services" 2013-2015 m.
  • Eurostars project "Production Effectiveness Navigator" 2011-2013.
  • Projects of EU Structural Funds LIEPA (Services Controlled by Lithuanian Voice), 2013-2015
  • Erasmus+ programme project "Partnerships to Ensure Risk Management in Practice" 2015-2016 m.
  • Projects of EU Structural Funds LIEPA2, 2017.12-2020.12
  • Postdoctoral Project “Analysis and Modeling of Promises Using the Joint Additive Synthesis and Source-Filter Model Method”, 2018-12-2019.12
  • COST Action CA15124 „A new Network of European BioImage Analysts to advance life science imaging (NEUBIAS)“. 2016-2020 m
  • COST Action CA15225 "Fractional-order systems - analysis, synthesis and their importance for future design", 2016-2020
  • COST Action CA18231 "Multi3Generation: Multi-task, Multilingual, Multi-modal Language Generation", 2019-2023.
  • National Competence Centres in the framework of EuroHPC, H2020, Project duration 2020-9-1 – 2022-8-31, Project Reference: 951732,
  • Postdoctoral Project “Investigating speech in the presence of noise interferences employing signal processing and machine learning methods”, 2021-07-2023.07