Tomasz Jach
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Tomasz Jach has a total of 14 co-authors. Here are the top co-authors by number of shared publications:
- Jan Kozak
University of Economics in Katowice
9 shared publications - Piotr Stefański
University of Economics in Katowice
9 shared publications - Barbara Probierz
University of Economics in Katowice
6 shared publications - Grzegorz Dziczkowski
Department of Machine Learning, University of Economics in Katowice, 1 Maja 50, 40-287 Katowice, Poland
5 shared publications - Przemysław Juszczuk
University of Economics in Katowice
4 shared publications - Szymon Głowania
Department of Machine Learning, University of Economics in Katowice, 1 Maja 50, 40-287 Katowice, Poland
4 shared publications
Latest Publications
- Open Access
- 18 February 2025
- Published by World Scientific Pub Co Pte Ltd in Vietnam Journal of Computer Science
- Vol. 13 (01) , 29-55
Machine Learning-Based Approach for the Gambling Problem Identification
Watching sports matches is a beloved pastime for every fan, and predicting the results and betting on which team will win adds to the excitement. Unfortunately, for some users, a lack of self-control can turn initially innocent play into a long-term problem. For this reason, problem gambling is among the crucial issues facing modern societies. Bookmaker companies, therefore, put considerable effort into implementing responsible gambling practices; however, identifying at-risk individuals remains a major challenge. Our paper focuses on identifying users who may currently have, or soon develop, potential issues with gambling. To achieve this, we introduce a set of machine learning methods combined with preprocessing tools that allow us to initially acquire and anonymize user data. This anonymized database is then used to identify high-risk groups. We tested our approach on a large dataset that was obtained and preprocessed specifically for this study. The experiments were conducted on actual data and verified by specialists responsible for identifying gambling problems. Using our method, we successfully identified and detected the early signs of potential gambling problems in multiple users.
- 18 February 2025
- Published by World Scientific Pub Co Pte Ltd in Vietnam Journal of Computer Science
- Vol. 13 (01) , 29-55
Open AccessMachine Learning-Based Approach for the Gambling Problem Identification
Watching sports matches is a beloved pastime for every fan, and predicting the results and betting on which team will win adds to the excitement. Unfortunately, for some users, a lack of self-control can turn initially innocent play into a long-term problem. For this reason, problem gambling is among the crucial issues facing modern societies. Bookmaker companies, therefore, put considerable effort into implementing responsible gambling practices; however, identifying at-risk individuals remains a major challenge. Our paper focuses on identifying users who may currently have, or soon develop, potential issues with gambling. To achieve this, we introduce a set of machine learning methods combined with preprocessing tools that allow us to initially acquire and anonymize user data. This anonymized database is then used to identify high-risk groups. We tested our approach on a large dataset that was obtained and preprocessed specifically for this study. The experiments were conducted on actual data and verified by specialists responsible for identifying gambling problems. Using our method, we successfully identified and detected the early signs of potential gambling problems in multiple users.
- Open Access
- 11 January 2025
- Published by MDPI AG in Applied Sciences
- Vol. 15 (2) , 667
Automated Tree Detection Using Image Processing and Multisource Data
This paper presents a method for the automatic detection and assessment of trees and tree-covered areas in Katowice, the capital of the Upper Silesian Industrial Region in southern Poland. The proposed approach utilizes satellite imagery and height maps, employing image-processing techniques and integrating data from various sources. We developed a data pipeline for gathering and pre-processing information, including vegetation data and numerical land-cover models, which were used to derive a new method for tree detection. Our findings confirm that automatic tree detection can significantly enhance the efficiency of urban tree management processes, contributing to the creation of greener and more resident-friendly cities.
- 11 January 2025
- Published by MDPI AG in Applied Sciences
- Vol. 15 (2) , 667
Open AccessAutomated Tree Detection Using Image Processing and Multisource Data
This paper presents a method for the automatic detection and assessment of trees and tree-covered areas in Katowice, the capital of the Upper Silesian Industrial Region in southern Poland. The proposed approach utilizes satellite imagery and height maps, employing image-processing techniques and integrating data from various sources. We developed a data pipeline for gathering and pre-processing information, including vegetation data and numerical land-cover models, which were used to derive a new method for tree detection. Our findings confirm that automatic tree detection can significantly enhance the efficiency of urban tree management processes, contributing to the creation of greener and more resident-friendly cities.
- 28 August 2024
- Published by Springer Nature
- p. 157-169
Improved CNN Model Stability and Robustness with Video Frame Segmentation
- 28 August 2024
- Published by Springer Nature
- p. 157-169
Improved CNN Model Stability and Robustness with Video Frame Segmentation
- Open Access
Boxing Punch Detection with Single Static Camera
Computer vision in sports analytics is gaining in popularity. Monitoring players’ performance using cameras is more flexible and does not interfere with player equipment compared to systems using sensors. This provides a wide set of opportunities for computer vision systems that help coaches, reporters, and audiences. This paper provides an introduction to the problem of measuring boxers’ performance, with a comprehensive survey of approaches in current science. The main goal of the paper is to provide a system to automatically detect punches in Olympic boxing using a single static camera. The authors use Euclidean distance to measure the distance between boxers and convolutional neural networks to classify footage frames. In order to improve classification performance, we provide and test three approaches to manipulating the images prior to fitting the classifier. The proposed solution achieves 95% balanced accuracy, 49% F1 score for frames with punches, and 97% for frames without punches. Finally, we present a working system for analyses of a boxing scene that marks boxers and labelled frames with detected clashes and punches.
Open AccessBoxing Punch Detection with Single Static Camera
Computer vision in sports analytics is gaining in popularity. Monitoring players’ performance using cameras is more flexible and does not interfere with player equipment compared to systems using sensors. This provides a wide set of opportunities for computer vision systems that help coaches, reporters, and audiences. This paper provides an introduction to the problem of measuring boxers’ performance, with a comprehensive survey of approaches in current science. The main goal of the paper is to provide a system to automatically detect punches in Olympic boxing using a single static camera. The authors use Euclidean distance to measure the distance between boxers and convolutional neural networks to classify footage frames. In order to improve classification performance, we provide and test three approaches to manipulating the images prior to fitting the classifier. The proposed solution achieves 95% balanced accuracy, 49% F1 score for frames with punches, and 97% for frames without punches. Finally, we present a working system for analyses of a boxing scene that marks boxers and labelled frames with detected clashes and punches.
- 16 July 2024
- Published by Springer Nature
- p. 263-274
Identification of Users in a Gambling Problem with the Use of Machine Learning
- 16 July 2024
- Published by Springer Nature
- p. 263-274
Identification of Users in a Gambling Problem with the Use of Machine Learning
- 13 September 2023
- Published by Springer Nature
- p. 540-551
Classification of Punches in Olympic Boxing Using Static RGB Cameras
- 13 September 2023
- Published by Springer Nature
- p. 540-551
Classification of Punches in Olympic Boxing Using Static RGB Cameras
- Open Access
- 1 January 2023
- Published by Elsevier in SSRN Electronic Journal
Boxing Punch Detection with Single Static Camera
Computer vision in sports analytics is gaining in popularity. Monitoring players' performance using cameras is more flexible and doesn't interfere with player equipment compared to systems using sensors. This provides a wide set of opportunities for computer vision systems that help coaches, reporters, and audiences. This paper provides an introduction to the problem of measuring boxers' performance in boxing, with a comprehensive research of approaches in current science. The main goal of the paper is to provide a system to automatically detect punches in Olympic boxing using a single static camera. The authors use Euclidean distance to measure the distance between boxers and Convolutional Neural Networks to classify footage frames. In order to improve classification performance we provide and test three approaches to manipulating the images prior to fitting the classifier. The proposed solution achieves 95% of balanced accuracy, 49% of F1 score for frames with punches, and 97% for frames without punches. Finally, we present a working system for analyses of the boxing scene that marks boxers, and labels frames with detected clashes and punches.
- 1 January 2023
- Published by Elsevier in SSRN Electronic Journal
Open AccessBoxing Punch Detection with Single Static Camera
Computer vision in sports analytics is gaining in popularity. Monitoring players' performance using cameras is more flexible and doesn't interfere with player equipment compared to systems using sensors. This provides a wide set of opportunities for computer vision systems that help coaches, reporters, and audiences. This paper provides an introduction to the problem of measuring boxers' performance in boxing, with a comprehensive research of approaches in current science. The main goal of the paper is to provide a system to automatically detect punches in Olympic boxing using a single static camera. The authors use Euclidean distance to measure the distance between boxers and Convolutional Neural Networks to classify footage frames. In order to improve classification performance we provide and test three approaches to manipulating the images prior to fitting the classifier. The proposed solution achieves 95% of balanced accuracy, 49% of F1 score for frames with punches, and 97% for frames without punches. Finally, we present a working system for analyses of the boxing scene that marks boxers, and labels frames with detected clashes and punches.
- 24 November 2022
- Published by Springer Nature
- p. 592-603
The Problem of Detecting Boxers in the Boxing Ring
- 24 November 2022
- Published by Springer Nature
- p. 592-603
The Problem of Detecting Boxers in the Boxing Ring
- Open Access
- 26 September 2022
- Published by Polish Information Processing Society PTI in Annals of Computer Science and Information Systems
- Vol. 30, 715-719
Modelling an IT solution to anonymise selected data processed in digital documents
- 26 September 2022
- Published by Polish Information Processing Society PTI in Annals of Computer Science and Information Systems
- Vol. 30, 715-719
Open AccessModelling an IT solution to anonymise selected data processed in digital documents