Institute of Applied Computer Science I-24
Head of the unit:
Anna Fabijańska, Ph.D., D.Sc., Assoc. Prof. Lodz University of Technology
Potential promoters:
Anna Fabijańska, Ph.D., D.Sc., Assoc. Prof. Lodz University of Technology
Contact person:
Anna Fabijańska, Ph.D., D.Sc., Assoc. Prof. Lodz University of Technology, tel: 42-631-27-50, anna.fabijańska@p.lodz.pl
Scope of activities:
Computer vision, digital image analysis and machine learning (including deep neural networks and graph neural networks) in selected problems in the field of medicine, industry and earth & environmental sciences.
Development of supervised and unsupervised methods of digital image segmentation.
Present activities:
- Development of methods for computer aided diagnosis in ophthalmology.
- Development of convolutional neural network models for computed aided dendrochronoligical and varve-based analysis (wood species recognition, tree-rings detection, resin ducts detection, automation of dendrochronological measurements, glacial varve (laminae) detection).
- Weak supervision in convolutional neural network based image segmentation.
- Automatic colorization of vintage movies using artificial intelligence methods.
Future activities:
Carry on the present activities listed above.
Publications/patents/awards/grants:
- Fabijańska A., Banasiak R.: Graph Convolutional Networks for Enhanced Resolution 3D Electrical Capacitance Tomography Image Reconstruction, Applied Soft Computing, vol. 110, 2021, 107608.
- Czepita M., Fabijańska A.: Image processing pipeline for the detection of blood flow through retinal vessels with subpixel accuracy in fundus images, Computer Methods and Programs in Biomedicine, vol. 208, 2021, 106240.
- Kucharski A., Fabijańska A.: CNN-Watershed: A Watershed Transform with Predicted Markers for Corneal Endothelium Image Segmentation, Biomedical Signal Processing and Control, vol. 68C, 2021, 102805.
- Affane A., Kucharski A., Chapuis P., Freydier S., Lebre M.A., Vacavant A., Fabijańska A.: Segmentation of liver anatomy by combining 3-D U-Net approaches, Applied Sciences, vol. 11, no. 11, 2021, str. 4895.
- Fabijańska A., Danek M.: Wood species automatic identification from wood core images with a residual convolutional neural network, Computers and Electronics in Agriculture, vol. 181C, 2021, str. 105941.
- Fabijańska A., Feder A., Rigde J.: DeepVarveNet: Automatic detection of glacial varves with deep neural networks, Computers & Geosciences, vol. 144, 2020, str. 104584.
- Fabijańska A., Grabowski S.: Viral Genome Deep Classifier, IEEE Access, vol. 7, 2019, str. 81297-81307.
- Chybicki M., Kozakiewicz W., Sielski D., Fabijańska A.: Deep cartoon colorizer: An automatic approach for colorization of vintage cartoons, Engineering Applications of Artificial Intelligence, vol. 81C, 2019, str. 37-46.
Keywords:
computer vision; image analysis; machine learning; neural networks; convolutional neural networks; image segmentation
The portfolio of research groups was created as part of the Programme "STER" – Internationalisation of doctoral schools” as part of the realization of the project “Curriculum for advanced doctoral education & taining – CADET Academy of Lodz University of Technology”.
Institute of Applied Computer Science I-24
Head of the unit:
Szymon Grabowski, PhD, DSc dr hab.
Potential promoters:
Szymon Grabowski, PhD, DSc dr hab.
Contact person:
Szymon Grabowski, PhD, DSc dr hab., tel. (+48) 42 631-27-50, Szymon.Grabowski@p.lodz.pl
Scope of activities:
- Design and implementation of bioinformatics tools, i.a., compressors for popular data types and formats (FASTQ, FASTA, genome collections).
- Efficient MEM (maximal exact matches) searching in bioinformatics data (pairs of genomes).
- Efficient algorithms for detecting similar sequences in collections.
- Efficient k-mer counting for sequencing data.
- Compact/succinct data structures (e.g., succinct representation for canonical Huffman code, with O(1)-time symbol encoding/decoding; an article under review, preprint: https://arxiv.org/abs/2108.05495).
Present activities:
- Development of a multiple bacteria genome compressor, MBGC (T.M.Kowalski, Sz.Grabowski).
- Development of efficient parallel MEM-finding algorithms (Sz.Grabowski, W.Bieniecki).
- Design and implementation of an efficient k-mer counting algorithm (Sz.Grabowski, W.Bieniecki, T.M.Kowalski, in collaboration with researchers from Ruđer Bošković Institute, Zagreb, Croatia).
- Design and implementation of an efficient FASTA compressor (Sz.Grabowski, R.Susik, T.M.Kowalski).
Future activities:
Following the topics being currently developed (see „present activities”). Development of the PgRC compressor of sequencing data (FASTA) (T.M.Kowalski, Szymon Grabowski, 2020), in collaboration with Piotr Duch. Design and implementation of indexing algorithms for repetitive data (i.a., with applications in bioinformatics). Finding similar sequences in collections (of, e.g., genomes) based on sketching.
Keywords:
text algorithms, bioinformatics, data compression
List of internship proposal in this research team:
- Compression of repetitive data (in particular, in bioinformatics), with random access.
- Indexing collections of repetitive data.
- Efficient algorithms for sequence similarity detection.
List of attachments:
Major publications of the team in 2019-2022:
- Szymon Grabowski, Tomasz Marek Kowalski: MBGC: Multiple Bacteria Genome Compressor. GigaScience, DOI:10.1093/gigascience/giab099, 2022 (accepted)
- Szymon Grabowski, Tomasz Marek Kowalski: Algorithms for all-pairs Hamming distance based similarity. Softw. Pract. Exp. 51(7): 1580-1590 (2021)
- Tomasz Marek Kowalski, Szymon Grabowski: PgRC: pseudogenome-based read compressor. Bioinform. 36(7): 2082-2089 (2020)
- Szymon Grabowski, Wojciech Bieniecki: copMEM: finding maximal exact matches via sampling both genomes. Bioinform. 35(4): 677-678 (2019)
- Tomasz Marek Kowalski, Szymon Grabowski, Kimmo Fredriksson: Suffix Arrays with a Twist. Comput. Informatics 38(3): 555- 574 (2019)
- Robert Susik, Szymon Grabowski, Kimmo Fredriksson: Revisiting Multiple Pattern Matching. Comput. Informatics 38(4): 937- 962 (2019)
The portfolio of research groups was created as part of the Programme "STER" – Internationalisation of doctoral schools” as part of the realization of the project “Curriculum for advanced doctoral education & taining – CADET Academy of Lodz University of Technology”.
Institute of Applied Computer Science I-24
Head of unit:
dr hab. inż. Andrzej Romanowski, univ. prof.
Potential promoters:
dr hab. inż. Andrzej Romanowski, univ. prof. dr hab. inż. Krzysztof Grudzień, univ. prof.
Contact person:
dr. inż. Magdalena Wróbel-Lachowska, tel: 42- 631-27-50, magdalena.wrobel-lachowska@p.lodz.pl
Scope of activities:
We operate in the interdisciplinary area of human-computer interaction; at the intersection of technical sciences (computer science, robotics, electronics, and artificial intelligence) with human and social sciences. We design, implement, and evaluate applications, systems and digital devices dedicated to cooperation with users within workplace, medical, office, industrial and everyday life environments. Our goal is to create and adapt emerging interactive technologies through a deep understanding of how modern technologies can support the needs of people and organizations. We conduct research and development in order to utilize the full potential of artificial intelligence for humans. We want to make the creation and use of digital technologies more useful, ergonomic, counteracting exclusion and consistent with the global goals of sustainable development. We use mixed methods in our research work; quantitative and qualitative, and we conduct our activities locally in the research initiative group of the Lodz University of Technology entitled ubicomp.pl, in cooperation with sigchi.pl, as well as within a wide international network of partners (LMU Munich, Chalmers Gothenburg, NUS Singapore, Harvard Boston and others).
Present activities:
Currently, a large part of our research work concerns supporting and enhancing the human senses & abilities as well as increasing the possibilities of developing the physical and mental potential of the digital systems users. We investigate ways of reducing the negative impact of human use of AI systems on the performance and development of users' cognitive skills (cooperation with Harvard, NUS and industrial partners). We create and develop an intelligent platform for remote support and telecare for lonely, dependent and people with special needs with the use of personal and wearable devices and machine learning methods (cooperation with HRP). We create devices and computer methods supporting athletes and amateurs (Chalmers, LMU), we create information processing and visualisation systems for industries of various industries using augmented and virtual reality (AR / VR) technology (Netrix, Oldenburg), IoT, and dedicated for Industry 4.0.
Future activities:
Research and development of environmental awareness systems in home, family, work, health care scenarios to support their own development, physical health and well-being. Design of interaction with AI systems, brain-computer interfaces (BCI) utilizing data gathered from crowdsourcing, oculographic and eye-tracking systems.
Publications/patents/awards/grants:
- To Trust or to Think: Cognitive Forcing Functions Can Reduce Overreliance on AI in AI-assisted Decision-making. PACMHCI 5, 2021, doi.org/10.1145/3449287
- Subtletee: Augmenting Posture Awareness for Beginner Golfers, in PACMHCI, ACM ISS, 2020. doi.org/10.1145/3427332 -
- Interactive Timeline Approach for Contextual Spatio-Temporal ECT Data Investigation Sensors, 20 (17), 4793, https://doi.org/10.3390/s20174793
- Considering Wake Gestures for Smart Assistant Use, 2020 CHI ACM, doi.org/10.1145/3334480.3383089
- Big Data-Driven Contextual Processing Methods for Electrical Capacitance Tomography, IEEE Transactions on Industrial Informatics, vol. 15, no. 3, pp. 1609-1618, 2019, doi: 10.1109/TII.2018.2855200.
- Using Crowdsourcing for Scientific Analysis of Industrial Tomographic Images. ACM Trans. Intell. Syst. Technol. 7, 4, Article 52 (2016), 25 p. DOI:https://doi.org/10.1145/2897370
- Curently we conduct several R&D projects and grants financed by national agencies and EU.
Keywords:
human-computer interaction, HCI, ubicomp, wearable computing, telecare systems
List of internship proposal in this research team:
Development of sensory systems, data collection, data processing methods, the use of machine learning, building data analysis and visualization platforms for a related network of users and stakeholders of intelligent interactive systems
The portfolio of research groups was created as part of the Programme "STER" – Internationalisation of doctoral schools” as part of the realization of the project “Curriculum for advanced doctoral education & taining – CADET Academy of Lodz University of Technology”.
Institute of Applied Computer Science I-24
Head of the unit:
Radosław Wajman, DSc. Ph.D. Eng., Prof. TUL
Scope of activities:
Among the area of my research the most important and the main research are the computer methods for non-invasive three-dimensional tomographic diagnostic and fuzzy control dedicated to the two-phase flow processes (TFP). TPFs arouse growing interest because of their great practical significance. They are closely related to the rapidly developing field of research in bioprocess engineering, biotechnology, environmental engineering, energy, and many other related branches. The growing needs of industry for simple, versatile, relatively inexpensive, non-invasive and rapid method of process diagnosis and control for TPFs in horizontal and vertical pipelines justify the importance of this topic. The knowledge of the characteristics and of the gas-liquid flow type is very important for the design and implementation of industrial-scale research facilities as well as for the process of numerical modelling. The continuous monitoring and diagnosis of any abnormalities can provide valuable information about their dynamic state and allow for continuous and automatic control. The scope of my research covered the development, implementation and verification of:
- raw tomographic measurement data processing algorithms in the context of TPF diagnosis;
- computer methods for spatial ECT sensor modelling and designing;
- fuzzy inference algorithms for the TPFs type identification and regulation;
- software deploying the developed algorithms and methods for the purpose of real flow processes monitoring and regulation.
Despite of the industrial purposes I develop methods of machine learning in medical applications. The main goal is to extract static and dynamic parameters in non-invasive diagnosis of lower urinary tracks.
Present activities:
- Computer Science in Medicine (algorithms of diagnostic data analysis and static/dynamic parameters identification)
- Computer Science for Industrial Applications (methods for non-invasive diagnosis and regulation of industrial processes)
Future activities:
Carry on the present activities listed above, with the emphasis on integrating AI methods for non-invasive diagnosis, monitoring and control.
Publications/patents/awards/grants:
- Fiderek, P., Kucharski, J., Wajman, R. (2021). Fuzzy Regulator for Two‐Phase Gas–Liquid Pipe Flows Control. Applied SciencesBasel, 1, 1-17. doi: 10.3390/app12010399
- Aghajanian, S., Rao, G., Ruuskanen, V., Wajman, R., Jackowska-Strumiłło, L., Koiranen, T. (2021). Real-Time Fault Detection and Diagnosis of CaCO3 Reactive Crystallization Process by Electrical Resistance Tomography Measurements. SENSORS, 21, 1-20.
- Wajman, R. (2021). The concept of 3D ECT system with increased border area sensitivity for crystallization processes diagnosis. Sensor Review, 1, 35-45. doi: 10.1108/SR-10-2019-0254
- Wajman, R. (2019). Computer methods for non-invasive measurement and control of two-phase flows: a review study. Information Technology and Control, 3, 464-486. doi: 10.5755/j01.itc.48.3.22189
The portfolio of research groups was created as part of the Programme "STER" – Internationalisation of doctoral schools” as part of the realization of the project “Curriculum for advanced doctoral education & taining – CADET Academy of Lodz University of Technology”.
Institute of Applied Computer Science I-24
Head of the unit:
Dr hab. Laurent Babout, univ. prof.
Potential promoters:
Dr hab. Laurent Babout, univ. prof.
Dr hab. inż. Robert Banasiak, univ. prof. Dr hab. inż. Krzysztof Grudzień, univ. prof.
Contact person:
Dr inż. Zbigniew Chaniecki, zbigniew.chaniecki@p.lodz.pl, (+48) 42 631-27-50 (w. 315)
Scope of activities:
The Institute of Applied Computer Science has a long track record of research activities in the field of process tomography, where the technology is used to support the diagnosis and monitoring of industrial processes linked to chemical engineering, food processing, raw materials handling to cite a few. We use our access to different modalities (electrical tomography, ultrasound tomography, X-ray tomography) and process installations to propose advanced algorithms to collect data via sensor development and to process and analyse it for different purposes (process visualisation, understanding, monitoring & control). We want to combine solid scientific foundation from applied mathematics and computer science (inverse problems, image reconstruction, image processing) with IT trends (AI/ML, VR/AR, parallel and distributed computing) to make the processed information more robust and easier to interpret or model. Our investigations tackle with development of modern, intelligent diagnostic platforms with an open architecture, meeting the expectations of Industry 4.0, with the possibility of free configuration and cooperation with external systems. We mainly conduct our activities at the Tom Dyakowski Process Tomography Lab (TDPTL) but also in world-class equipment via cooperation with national and international strategic partners (netrix.pl, HZDR (Germany), INSALyon (France), TU Delft (The Netherlands)).
Present activities:
We currently address a variety of scientific challenges linked to process tomography and related field:
- Process control of gas-water separation efficiency using electrical resistance tomography – focus on alternative methods to image reconstruction for realtime control (MSCA project TOMOCON)
- Real-time overlay of tomography data on corresponding process using handheld-based augmented reality (activity annex to TOMOCON)
- Enhancement of image reconstruction of 3D ECT using deep learning
- Fast tomography data processing using massive parallel and distributed computations
- Development of image processing methods to analyse raw material properties (organic materials, meteorites) extracted from X-ray tomography images
Future activities:
Carry on the present activities listed above, with the emphasis on integrating IT trends for optimized real-time/online monitoring and diagnosis.
Publications/patents/awards/grants:
- Graph convolutional networks for enhanced resolution 3D Electrical Capacitance Tomography image reconstruction, Applied Soft Computing (2021), 110: 107608, https://doi.org/10.1016/j.asoc.2021.107608
- Multichannel Capacitive Imaging of Gas Vortex in Swirling Two-Phase Flows Using Parametric Reconstruction. IEEE Access (2020), 8: 69557-69565, https://doi.org/10.1109/ACCESS.2020.2986724
- On the Use of a Rotatable ECT Sensor to Investigate Dense Phase Flow: A Feasibility Study. Sensors (2020), 20: 4854, https://doi.org/10.3390/s20174854
- Quantitative analysis of flow dynamics of organic granular materials inside a versatile silo model during time-lapse X-ray tomography experiments. Computers and Electronics in Agriculture (2020), 172: 105346, https://doi.org/10.1016/j.compag.2020.105346
- Analysis of silo flow dynamic effects using ECT and short time Fourier transform. Flow Measurement and Instrumentation (2018), 62: 167-175, https://doi.org/ 10.1016/j.flowmeasinst.2018.02.003
- https://www.tomocon.eu
Keywords:
Process visualization; image processing; data handling; 3D printing; sensor design; tomography modalities for applied science
List of internship proposal in this research team:
Development of tomography sensors, 3D modelling and printing, data collection schemes, data handling methods, the use of machine learning and artificial intelligence, parallel and distributed computing
The portfolio of research groups was created as part of the Programme "STER" – Internationalisation of doctoral schools” as part of the realization of the project “Curriculum for advanced doctoral education & taining – CADET Academy of Lodz University of Technology”.