00338nas a2200145 4500000000100000000000100001008004100002100001900043700002100062700002000083700001900103700002400122245003900146490000700185 2020 d1 aAdam Domański1 aJoanna Domańska1 aKatarzyna Filus1 aJakub Szyguła1 aTadeusz Czachórski00aThe self-similar markovian sources0 v1000430nas a2200157 4500000000100000000000100001008004100002260000900043100002000052700001900072700002100091700001800112700001900130245011600149490000700265 2020 d c20201 aKatarzyna Filus1 aAdam Domański1 aJoanna Domańska1 aDariusz Marek1 aJakub Szyguła00aLong-Range Dependent Traffic Classification with Convolutional Neural Networks Based on Hurst Exponent Analysis0 v2200540nas a2200181 4500000000100000000000100001008004100002260003800043100002000081700002200101700002100123700001700144245008700161856006600248300001200314490001000326022002200336 2021 d bSpringer International Publishing1 aKatarzyna Filus1 aMiltiadis Siavvas1 aJoanna Domańska1 aErol Gelenbe00aThe Random Neural Network as a Bonding Model for Software Vulnerability Prediction uhttps://link.springer.com/chapter/10.1007/978-3-030-68110-4_7 a102-1160 v12527 a978-3-030-68110-401179nas a2200181 4500000000100000000000100001008004100002260003800043100002000081700002100101700001700122245007000139856006600209300001000275490001000285520068000295022002200975 2021 d bSpringer International Publishing1 aKatarzyna Filus1 aJoanna Domańska1 aErol Gelenbe00aRandom Neural Network for Lightweight Attack Detection in the IoT uhttps://link.springer.com/chapter/10.1007/978-3-030-68110-4_5 a79-910 v125273 a

Cyber-attack detection has become a basic component of all information processing systems, and once an attack is detected it may be possible to block or mitigate its effects. This paper addresses the use of a learning recurrent Random Neural Network (RNN) to build a lightweight detector for certain types of Botnet attacks on IoT systems. Its low computational cost based on a small 12-neuron recurrent architecture makes it particularly attractive for edge devices. The RNN can be trained off-line using a fast simplified gradient descent algorithm, and we show that it can lead to high detection rates of the order of 96%, with false alarm rates of a few percent.

 a978-3-030-68110-400460nas a2200169 4500000000100000000000100001008004100002100002000043700002000063700002100083700002200104700001700126245007700143856004500220490001100265022001400276 2021 d1 aKatarzyna Filus1 aPaweł Boryszko1 aJoanna Domańska1 aMiltiadis Siavvas1 aErol Gelenbe00aEfficient Feature Selection for Static Analysis Vulnerability Prediction uhttps://www.mdpi.com/1424-8220/21/4/11330 v21 (4) a1424-822000416nas a2200157 4500000000100000000000100001008004100002100002000043700002000063700001900083700002100102245006300123856004600186490001200232022001400244 2021 d1 aŁukasz Sobczak1 aKatarzyna Filus1 aAdam Domański1 aJoanna Domańska00aLIDAR Point Cloud generation for SLAM algorithm evaluation uhttps://www.mdpi.com/1424-8220/21/10/33130 v21 (10) a1424-822000473nas a2200169 4500000000100000000000100001008004100002100001900043700001900062700002100081700001800102700002000120700001800140245008800158856004600246490001100292 2021 d1 aJakub Szyguła1 aAdam Domański1 aJoanna Domańska1 aDariusz Marek1 aKatarzyna Filus1 aSzymon Mendla00aSupervised learning of Neural Networks for Active Queue Management in the Internet uhttps://www.mdpi.com/1424-8220/21/15/49790 v21(15)16164nas a2200253 45000000001000000000001000010080041000022600021000431000025000647000022000897000024001117000023001357000025001587000020001837000020002037000021002237000026002447000029002707000017002997000023003167000022003392450090003615201545900451 2022 d c03/2022aBelgium1 aCharalampos Marantos1 aMiltiadis Siavvas1 aDimitrios Tsoukalas1 aChristos Lamprakos1 aLazaros Papadopoulos1 aPaweł Boryszko1 aKatarzyna Filus1 aJoanna Domańska1 aApostolos Ampatzoglou1 aAlexander Chatzigeorgiou1 aErol Gelenbe1 aDionysios Kehagias1 aDimitrios Soudris00aSDK4ED: One-click platform for Energy-aware, Maintainable and Dependable Applications3 a

Developing modern secure and low-energy applications in a short time imposes new challenges and creates the need of designing new software tools to assist developers in all phases of application development. The design of such tools cannot be considered a trivial task, as they should be able to provide optimization of multiple quality requirements. In this paper, we introduce the SDK4ED platform, which incorporates advanced methods and tools for measuring and optimizing maintainability, dependability and energy. The presented solution offers a complete tool-flow for providing indicators and optimization methods with emphasis on embedded software. Effective forecasting models and decision-making solutions are also implemented to improve the quality of the software, respecting the constraints imposed on maintenance standards, energy consumption limits and security vulnerabilities. The use of the SDK4ED platform is demonstrated in a healthcare embedded application.

00451nas a2200145 4500000000100000000000100001008004100002100002000043700002000063700002100083700001500104245012400119856005500243490000700298 2022 d1 aKatarzyna Filus1 aSławomir Nowak1 aJoanna Domańska1 aJakub Duda00aCost-Effective Filtering of Unreliable Proximity Detection Results Based on BLE RSSI and IMU Readings Using Smartphones uhttps://www.nature.com/articles/s41598-022-06201-y0 v1200461nas a2200181 4500000000100000000000100001008004100002100001800043700001900061700001900080700002100099700002000120700002100140245005300161856004400214490000700258022001400265 2022 d1 aDariusz Marek1 aJakub Szyguła1 aAdam Domański1 aJoanna Domańska1 aKatarzyna Filus1 aMarta Szczygieł00aAdaptive Hurst-Sensitive Active Queue Management uhttps://www.mdpi.com/1099-4300/24/3/4180 v24 a1099-430000340nas a2200121 4500000000100000000000100001008004100002260003500043100002000078700002100098245004100119856005800160 2022 d bIEEEaPadova, 18-23 July, 20221 aKatarzyna Filus1 aJoanna Domańska00aNAM: What Does a Neural Network See? uhttps://ieeexplore.ieee.org/abstract/document/989244200511nas a2200181 4500000000100000000000100001008004100002100001800043700001900061700002100080700001900101700002400120700001700144700002000161245006400181856007700245490000700322 2022 d1 aDariusz Marek1 aAdam Domański1 aJoanna Domańska1 aJakub Szyguła1 aTadeusz Czachórski1 aJerzy Klamka1 aKatarzyna Filus00aApproximation Models for the Evaluation of TCP/AQM Networks uhttps://journals.pan.pl/dlibra/publication/141986/edition/123961/content0 v7000397nas a2200133 4500000000100000000000100001008004100002260001200043100002000055700002100075245007600096856008300172490000800255 2023 d c10/20221 aKatarzyna Filus1 aJoanna Domańska00aSoftware Vulnerabilities in TensorFlow-Based Deep Learning Applications uhttps://www.sciencedirect.com/science/article/pii/S0167404822003406?via%3Dihub0 v12400454nas a2200145 4500000000100000000000100001008004100002100002000043700002000063700002100083700001900104245012300123856005500246490000700301 2022 d1 aŁukasz Sobczak1 aKatarzyna Filus1 aJoanna Domańska1 aAdam Domański00aFinding the best hardware configuration for 2D SLAM in indoor environments via simulation based on Google Cartographer uhttps://www.nature.com/articles/s41598-022-22938-y0 v1200409nas a2200145 4500000000100000000000100001008004100002100002000043700002000063700002100083700001900104245008700123856004600210490000700256 2022 d1 aŁukasz Sobczak1 aKatarzyna Filus1 aJoanna Domańska1 aAdam Domański00aBuilding a real-time testing platform for unmanned ground vehicles with UDP Bridge uhttps://www.mdpi.com/1424-8220/22/21/84930 v2200460nas a2200157 4500000000100000000000100001008004100002260003800043100002000081700002000101700002100121700001900142700001700161245007400178856005000252 2022 d bIEEEaOsaka, 17-20 December, 20221 aKatarzyna Filus1 aŁukasz Sobczak1 aJoanna Domańska1 aAdam Domański1 aRafał Cupek00aReal-time testing of vision-based systems for AGVs with ArUco markers uhttps://ieeexplore.ieee.org/document/1002081700353nas a2200121 4500000000100000000000100001008004100002100002000043700002100063245006700084856007200151490000800223 2023 d1 aKatarzyna Filus1 aJoanna Domańska00aGlobal Entropy Pooling Layer for Convolutional Neural Networks uhttps://www.sciencedirect.com/science/article/pii/S09252312230073850 v55500302nas a2200109 4500000000100000000000100001008004100002260003100043100002000074700002100094245007700115 2023 d bIEEEaThessaloniki, Greece1 aKatarzyna Filus1 aJoanna Domańska00aRecycling of Generic ImageNet-trained Models for Smart-city Applications01730nas a2200145 4500000000100000000000100001008004100002260003600043100001800079700002000097700002100117245010300138856005000241520129300291 2023 d c01/2024bIEEEaSorrento, Italy 1 aNur Kelesoglu1 aKatarzyna Filus1 aJoanna Domańska00aHierAct: a Hierarchical Model for Human Activity Recognition in Game-Like Educational Applications uhttps://ieeexplore.ieee.org/document/103869193 a

Constantly evolving landscape of modern education makes the integration of technology crucial to enable more interactive and immersive learning experiences. One of the key problems in this domain is Human Activity Recognition (HAR), which uses standard smartphone sensors to understand and recognize user movements. While HAR exhibits high potential in this area, its persistent challenge is to differentiating between similar activities, whose occurrences can often lead to misclassification and reduced quality of applications. To tackle this problem, we employ specialized classifiers and use them in a hierarchical manner instead of using monolith multi-class models. We propose HierAct - a Hierarchical model that solves the multi-class human activity recognition problem. We also release a new HAR dataset - EduAct - (https://github.com/iitis/HierAct-Dataset) focused on activities that could be used to create game-like educational applications.Our results show that the performance of the proposed model is substantially better than other state-of-the-art machine learning models. Our work shows the potential of hierarchical classifiers in HAR applications for education, by offering educators and students a more accurate, reliable, and engaging interactive experience.

00389nas a2200121 4500000000100000000000100001008004100002260007200043100001900115700002000134700002100154245009200175 2023 d c01/2024bIEEE International Conference on Big DataaSorrento, Italy1 aMarzena Halama1 aKatarzyna Filus1 aJoanna Domańska00aRobust category recognition based on deep templates for educational mobile applications00264nas a2200109 4500000000100000000000100001008004100002260002000043100002000063700002100083245005000104 2023 d aSorrento, Italy1 aKatarzyna Filus1 aJoanna Domańska00aNetSat: Network Saturation Adversarial Attack00387nas a2200133 4500000000100000000000100001008004100002260003500043100002000078700002000098700001900118700002100137245009500158 2023 d c01/2024bIEEEaSorrento, Italy1 aŁukasz Sobczak1 aKatarzyna Filus1 aMarzena Halama1 aJoanna Domańska00aVisual examination of relations between known classes for deep neural network classifiers.00838nas a2200289 4500000000100000000000100001008004100002100002200043700002400065700002500089700002500114700002300139700001700162700002100179700002100200700002600221700002000247700002100267700001900288700002600307700002200333700002900355700001700384700002300401700002300424245010100447 2024 d1 aMiltiadis Siavvas1 aDimitrios Tsoukalas1 aCharalambos Marantos1 aLazaros Papadopoulos1 aChristos Lamprakos1 aOliviu Matei1 aChristos Strydis1 aMuhammad Siddiqi1 aPhilippe Chrobocinski1 aKatarzyna Filus1 aJoanna Domańska1 aParis Avgeriou1 aApostolos Ampatzoglou1 aDimitrios Soudris1 aAlexander Chatzigeorgiou1 aErol Gelenbe1 aDionysios Kehagias1 aDimitrios Tzovaras00aSDK4ED: a platform for building energy efficient, dependable, and maintainable embedded software00447nas a2200133 4500000000100000000000100001008004100002100002800043700002200071700002700093700002000120245012300140856005000263 2024 d1 aGustavo Machado-Guillen1 aJorge Cruz-Duarte1 aSantiago Conant-Pablos1 aKatarzyna Filus00aHarnessing Machine Learning for Reliable Weather Forecasting: Meteorological Impact on Sustainable Energy in Monterrey uhttps://ieeexplore.ieee.org/document/1061199300417nas a2200121 4500000000100000000000100001008004100002100002600043700002000069700001900089245011500108856007200223 2024 d1 aSebastian Pokuciński1 aKatarzyna Filus1 aDariusz Mrozek00aFinding the Perfect Cut: Selection of the Best Cutting Point in Equirectangular Panoramas for Object Detection uhttps://www.sciencedirect.com/science/article/pii/S187705092402472400393nas a2200133 4500000000100000000000100001008004100002260001200043100002000055700002100075245008300096856007200179490000800251 2024 d c12/20241 aKatarzyna Filus1 aJoanna Domańska00aExtracting coarse-grained classifiers from large Convolutional Neural Networks uhttps://www.sciencedirect.com/science/article/pii/S09521976240153550 v13800328nas a2200109 4500000000100000000000100001008004100002100002000043700002100063700001700084245011700101 2024 d1 aKatarzyna Filus1 aJoanna Domańska1 aJerzy Klamka00aEvaluating Convolutional Neural Networks via measuring similarity of their first-layer filters and Gabor filters00371nas a2200133 4500000000100000000000100001008004100002260001200043100002000055700002100075245006100096856007200157490000800229 2024 d c12/20241 aKatarzyna Filus1 aJoanna Domańska00aSimilarity-driven adversarial testing of neural networks uhttps://www.sciencedirect.com/science/article/pii/S09507051240125530 v30500323nas a2200121 4500000000100000000000100001008004100002100002000043700002100063245009500084490000800179022001400187 2025 d1 aKatarzyna Filus1 aJoanna Domańska00aWhat is the doggest dog? Examination of typicality perception in ImageNet-trained networks0 v188 a0893-608000357nas a2200121 4500000000100000000000100001008004100002100002000043700002400063700002000087245010700107856002100214 d1 aKatarzyna Filus1 aMichał Romaszewski1 aMateusz Żarski00aSemantic Depth Matters: Explaining Errors of Deep Vision Networks through Perceived Class Similarities uarXiv:2504.0995601423nas a2200121 4500000000100000000000100001008004100002100002000043700002000063245012200083856003700205520105900242 d1 aKatarzyna Filus1 aMateusz Żarski00aBeyond Accuracy: Uncovering the Role of Similarity Perception and its Alignment with Semantics in Supervised Learning uhttps://arxiv.org/abs/2505.213383 a

Similarity manifests in various forms, including semantic similarity that is particularly important, serving as an approximation of human object categorization based on e.g. shared functionalities and evolutionary traits. It also offers practical advantages in computational modeling via lexical structures such as WordNet with constant and interpretable similarity. As in the domain of deep vision, there is still not enough focus on the phenomena regarding the similarity perception emergence. We introduce Deep Similarity Inspector (DSI) -- a systematic framework to inspect how deep vision networks develop their similarity perception and its alignment with semantic similarity. Our experiments show that both Convolutional Neural Networks' (CNNs) and Vision Transformers' (ViTs) develop a rich similarity perception during training with 3 phases (initial similarity surge, refinement, stabilization), with clear differences between CNNs and ViTs. Besides the gradual mistakes elimination, the mistakes refinement phenomenon can be observed.

00323nas a2200109 4500000000100000000000100001008004100002100002000043700002100063245006300084856006600147 2025 d1 aKatarzyna Filus1 aJoanna Domańska00aHow CNNs and ViTs perceive similarities between categories uhttps://link.springer.com/chapter/10.1007/978-3-032-06078-5_200396nas a2200121 4500000000100000000000100001008004100002260000900043100002000052700002600072245009300098856008300191 2025 d c20251 aKatarzyna Filus1 aSebastian Pokuciński00aSPHEREA: Synthetic dataset generation for equirectangular panoramas of home environments uhttps://www.sciencedirect.com/science/article/pii/S0950705125017897?via%3Dihub00257nas a2200085 4500000000100000008004100001100002000042700002200062245008700084 2026 d1 aKatarzyna Filus1 aJorge Cruz-Duarte00aTarget Selection for Adversarial Evaluation of Vision Models Using Language Models00536nas a2200157 4500000000100000008004100001100002000042700002200062700002000084700002300104700002000127700002100147700002100168245012200189856006700311 2026 d1 aKatarzyna Filus1 aJorge Cruz-Duarte1 aMateusz Korpyś1 aAndrzej Kołodziej1 aAnna Gancarczyk1 aEl-Ghazali Talbi1 aJoanna Domańska00aFeature- and Similarity-Based Explainability of Experimental Materials Data via Black-Box Hyperparameter Optimization uhttps://link.springer.com/chapter/10.1007/978-3-032-36226-1_1100258nas a2200085 4500000000100000008004100001100002000042700002000062245009000082 2026 d1 aKatarzyna Filus1 aMateusz Żarski00aInspecting Training Dynamics of Similarity Development in Supervised Vision Networks 00259nas a2200085 4500000000100000008004100001100002000042700002600062245008500088 2026 d1 aKatarzyna Filus1 aSebastian Pokuciński00aMeasuring Monosemanticity in Sparse Autoencoders via Latent Activation Coherence00381nas a2200121 4500000000100000008004100001100002000042700002000062700001800082700002100100700003000121245010800151 2026 d1 aKatarzyna Filus1 aObinna Nwadiuto1 aZlatka Stoeva1 aJoanna Domańska1 aFideline Tchuenbou-Magaia00aLightweight Time-Series Representation for Battery State of Health Estimation from Graphene Sensor Data00310nas a2200109 4500000000100000008004100001100002000042700001600062700002000078700002200098245008000120 2026 d1 aKatarzyna Filus1 aKamil Faber1 aRoberto Corizzo1 aChristopher Kanan00aLost or Hidden? A Concept-Level Forgetting in Supervised Continual Learning