A5010915319- Astrophysics and Cosmic Phenomena
- Dark Matter and Cosmic Phenomena
- Particle Detector Development and Performance
- Human Pose and Action Recognition
- Handwritten Text Recognition Techniques
- Hand Gesture Recognition Systems
- Anomaly Detection Techniques and Applications
- User Authentication and Security Systems
- Cognitive Computing and Networks
- Video Analysis and Summarization
- Natural Language Processing Techniques
- Martial Arts: Techniques, Psychology, and Education
- Neutrino Physics Research
- Sensor Technology and Measurement Systems
- Biometric Identification and Security
- Sports Dynamics and Biomechanics
- CCD and CMOS Imaging Sensors
- Earthquake Detection and Analysis
- Computational Physics and Python Applications
- Time Series Analysis and Forecasting
- Gamma-ray bursts and supernovae
- Solar and Space Plasma Dynamics
- Multidisciplinary Science and Engineering Research
- Human Motion and Animation
- Banking Systems and Strategies
Jagiellonian University
2023-2025
AGH University of Krakow
2021-2025
Institute of Nuclear Physics, Polish Academy of Sciences
2021-2023
Uniwersytet Komisji Edukacji Narodowej w Krakowie
2014-2023
Institute of Computer Science
2012-2023
Cracow University of Technology
2021
Pedagogical University
2014
The aim of this paper is to propose and evaluate the novel method template generation, matching, comparing visualization applied motion capture (kinematic) analysis. To our approach, we have used recordings (MoCap) two highly-skilled black belt karate athletes consisting 560 various techniques acquired with wearable sensors. We evaluated quality generated templates; validated matching algorithm that calculates similarities differences between MoCap data; examined visualizations important...
The Cosmic Ray Extremely Distributed Observatory (CREDO) is a newly formed, global collaboration dedicated to observing and studying cosmic rays (CR) ray ensembles (CRE): groups of minimum two CR with common primary interaction vertex or the same parent particle. CREDO program embraces testing known CRE scenarios, preparing observe unexpected physics, it also suitable for multi-messenger multi-mission applications. Perfectly matched capabilities, could be formed both within classical models...
In this paper, we address the issues of explainability reinforcement learning-based machine learning agents trained with Proximal Policy Optimization (PPO) that utilizes visual sensor data. We propose an algorithm allows effective and intuitive approximation PPO-trained neural network (NN). conduct several experiments to confirm our method’s effectiveness. Our proposed method works well for scenarios where semantic clustering scene is possible. approach based on solid theoretical foundation...
Abstract Propagation of ultra-high energy photons in the solar magnetosphere gives rise to cascades comprising thousands photons. We study cascade development using Monte Carlo simulations and find that are spatially extended over millions kilometers on plane distant from Sun by 1 AU. estimate chance detection considering upper limits current cosmic rays observatories order provide an optimistic rate 0.002 events per year a chosen ring-shaped region around Sun. compare results which use two...
We present the purpose, long-term development vision, basic design, detection algorithm and preliminary results obtained with Cosmic Ray Extremely Distributed Observatory (CREDO) Detector mobile application. The CREDO app related infrastructure are unique in terms of their scale, targeting many form-factors open-access philosophy. This philosophy translates to open-source code app, both data inflow as well consumption above all, citizen science that means is open all who wish participate...
The motivation of this paper is to examine the effectiveness state-of-the-art and newly proposed motion capture pattern recognition methods in task head gesture classifications. gestures are designed for a user interface that utilizes virtual reality helmet equipped with an internal measurement unit (IMU) sensor has 6-axis accelerometer gyroscope. We will validate classifier uses Principal Components Analysis (PCA)-based features various numbers dimensions, two-stage PCA-based method,...
Gamification is known to enhance users’ participation in education and research projects that follow the citizen science paradigm. The Cosmic Ray Extremely Distributed Observatory (CREDO) experiment designed for large-scale study of various radiation forms continuously reach Earth from space, collectively as cosmic rays. CREDO Detector app relies on a network involved users now working worldwide across phones other CMOS sensor-equipped devices. To broaden user base activate current users,...
Applying Machine Learning (ML) methods for the analysis of muon lateral distributions in Extensive Air Showers detected by citizen science projects, while taking into account spatial distribution detectors requires enormous training data sets. Therefore, generating these sets with typical Monte Carlo (MC) generators like CORSIKA is computationally prohibitive. Here we present a method which application special augmentation procedures produces dataset that compatible all essential aspects to...
The article presents the approach based on usage of syntactic methods for static analysis handwritten signatures. graph linguistic formalisms applied, such as IE and ETPL(k) grammar, are characterised by considerable descriptive strength a polynomial membership problem analysis. For purposes representing analysed signatures, new hierarchical (two-layer) HIE structures graphs have been defined. two-layer description makes it possible to take into consideration both local global features...
The last decade has brought about a profound transformation in multimessenger science. Ten years ago, facilities had been built or were under construction that would eventually discover the nature of objects our universe could be detected through multiple messengers. Nonetheless, science was hardly more than dream. rewards for foresight finally realized IceCube's discovery diffuse astrophysical neutrino flux, first observation gravitational waves by LIGO, and joint detections photons...
In this paper, we discuss a practice of potential cosmic ray detection using off-the-shelves CMOS cameras. We and presents the limitations up-to-date hardware software approaches to task. also present solution that made for long-term testing algorithms detection. have proposed, implemented tested novel algorithm enables real-time processing image frames acquired by cameras in order detect tracks particles. compared our results with already published obtained acceptable overcoming some...
In this paper, we propose using the radial basis functions (RBF) to determine upper bound of absolute dynamic error (UAE) at output a voltage-mode accelerometer. Such can be obtained as result approximating values determined for assumed-in-advance parameter variability associated with mathematical model an This approximation was carried out function neural network (RBF-NN) procedure given number neurons. The Monte Carlo (MC) method also applied related when considering uncertainties...
In this paper, we describe the convolutional neural network (CNN)-based approach to problems of categorization and artefact reduction cosmic ray images obtained from CMOS sensors used in mobile phones. As artefacts, understand all that cannot be attributed particles’ passage through sensor but rather result deficiencies registration procedure. The proposed deep is composed a pretrained CNN neural-network-based approximator, which models uncertainty image class assignment. was trained using...
In this paper we propose the method for detecting potential anomalous cosmic ray particle tracks in big data image dataset acquired by Complementary Metal-Oxide-Semiconductors (CMOS). Those sensors are part of scientific infrastructure Cosmic Ray Extremely Distributed Observatory (CREDO). The use Incremental PCA (Principal Components Analysis) allowed approximation loadings which might be updated at runtime. with Sequential Karhunen-Loeve Transform results almost identical embedding as basic...
The teaching of motion activities in rehabilitation, sports, and professional work has great social significance. However, the automatic these activities, particularly those involving fast motions, requires use an adaptive system that can adequately react to changing stages conditions process. This paper describes a prototype utilizes online classification signals select proper algorithm. knowledge necessary perform process is acquired from experts by machine learning methodology....
Reliable tools for artefact rejection and signal classification are a must cosmic ray detection experiments based on CMOS technology. In this paper, we analyse the fitness of several feature-based statistical classifiers particle candidate hits in four categories: spots, tracks, worms artefacts. We use Zernike moments image function as feature carriers propose preprocessing denoising scheme to make extraction more efficient. As opposed convolution neural network classifiers, allow...
The Cosmic Ray Extremely Distributed Observatory (CREDO) pursues a global research strategy dedicated to the search for correlated cosmic rays, so-called Ensembles (CRE). Its general approach CRE detection does not involve any priori considerations, and its encompasses both spatial temporal correlations, on different scales. Here we time clustering of ray events collected with small sea-level extensive air shower array at University Adelaide. consists seven one-square-metre scintillators...
In this paper, a graph-based off-line handwritten signature verification system is proposed. The can automatically identify some global and local features which exist within different signatures of the same person. Based on these it possible to verify whether forgery or not. structural description in form hierarchical attributed random graph set transformed into matrix-vector structures. These structures be directly used as matching pattern when examined analyzed. proposed approach applied...
The Frangi neuron proposed in this work is a complex element that allows high-level Hessian-based image processing. Its adaptive parameters (weights) can be trained using minimum number of training data. In our experiment, we showed just one enough to optimize the values weights. An intuitive application use it segmentation process. order test performance neuron, used diverse medical datasets on which second-order structures are visualized. network presented paper single proved significantly...
In this paper we consider possibility of using palm movements as effective behavioral biometric modality. We propose to exploit the hand motion characteristics gathered from 3D sensor device input data for identification system. Currently various reasons people are concerned about directly touching scanners, therefore touch-less and non-invasive approach seems be useful in practical applications. described scheme fingertip positions tracked real-time, while defined gesture is performed. The...