<?xml version="1.0" encoding="UTF-8"?>
<xml><records><record><source-app name="Bibcite" version="8.x">Drupal-Bibcite</source-app><ref-type>17</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Mateusz Żarski</style></author><author><style face="normal" font="default" size="100%">Bartosz Wójcik</style></author><author><style face="normal" font="default" size="100%">Kamil Książek</style></author><author><style face="normal" font="default" size="100%">Jaroslaw Miszczak</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">Finicky transfer learning—A method of pruning convolutional neural networks for cracks classification on edge devices</style></title></titles><keywords/><dates><year><style face="normal" font="default" size="100%">2021</style></year><pub-dates><date><style face="normal" font="default" size="100%">03/2022</style></date></pub-dates></dates><secondary-title><style face="normal" font="default" size="100%">Computer-Aided Civil And Infrastructure Engineering</style></secondary-title><electronic-resource-num><style face="normal" font="default" size="100%">10.1111/mice.12755</style></electronic-resource-num><volume><style face="normal" font="default" size="100%">37</style></volume><issue><style face="normal" font="default" size="100%">4</style></issue><section><style face="normal" font="default" size="100%">500</style></section><abstract><style face="normal" font="default" size="100%">&lt;p&gt;High demand for computational power significantly limits the possibility of using modern deep learning methods in the environments where one has to deal with devices limited by the performance and the energy constraints. To address this issue, this paper proposes a novel method of combining the pruning and the transfer learning techniques for the purpose of delivering solid accuracy while simultaneously lowering the demand for energy and computing power. This method is referred to as \emph{Finicky Transfer Learning} (FTL) as it is finicky during the process of selecting filters from a pre-trained feature extractor to compose a sparser architecture. The proposed filter selection process is based on an original approach utilizing the Jaccard similarity coefficient calculated between the activation maps and the masks obtained by semantic segmentation. This enables the use of CNN, trained previously on a large generic dataset, in a crack classification task. The presented method significantly lowers the inference time while maintaining or even slightly increasing the classification accuracy, enabling real-time operation on single-board computers.&lt;/p&gt;&#13;
</style></abstract><language><style face="normal" font="default" size="100%">eng</style></language></record><record><source-app name="Bibcite" version="8.x">Drupal-Bibcite</source-app><ref-type>17</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Bartosz Wójcik</style></author><author><style face="normal" font="default" size="100%">Mateusz Żarski</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">The measurements of surface defect area with an RGB-D camera for a BIM-backed bridge inspection</style></title></titles><keywords/><dates><year><style face="normal" font="default" size="100%">2021</style></year><pub-dates><date><style face="normal" font="default" size="100%">6/2021</style></date></pub-dates></dates><secondary-title><style face="normal" font="default" size="100%">Bulletin of the Polish Academy of Sciences: Technical Sciences</style></secondary-title><electronic-resource-num><style face="normal" font="default" size="100%">10.24425/bpasts.2021.137123</style></electronic-resource-num><volume><style face="normal" font="default" size="100%">69</style></volume><issue><style face="normal" font="default" size="100%">3</style></issue><language><style face="normal" font="default" size="100%">eng</style></language></record><record><source-app name="Bibcite" version="8.x">Drupal-Bibcite</source-app><ref-type>17</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Mateusz Żarski</style></author><author><style face="normal" font="default" size="100%">Bartosz Wójcik</style></author><author><style face="normal" font="default" size="100%">Jaroslaw Miszczak</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">KrakN: Transfer Learning framework and dataset for infrastructure thin crack detection</style></title></titles><keywords/><dates><year><style face="normal" font="default" size="100%">2021</style></year><pub-dates><date><style face="normal" font="default" size="100%">12/2021</style></date></pub-dates></dates><secondary-title><style face="normal" font="default" size="100%">SoftwareX</style></secondary-title><urls><style face="normal" font="default" size="100%">https://arxiv.org/abs/2004.12337</style></urls><electronic-resource-num><style face="normal" font="default" size="100%">10.1016/j.softx.2021.100893</style></electronic-resource-num><volume><style face="normal" font="default" size="100%">16</style></volume><section><style face="normal" font="default" size="100%">100893</style></section><abstract><style face="normal" font="default" size="100%">&lt;p&gt;Monitoring the technical condition of infrastructure is a crucial element of its maintenance. Although there are many deep learning models intended for this purpose, they are severely limited in their application due to labour-intensive gathering of new datasets and high demand for computing power during model training. To overcome these limiting factors we propose a KrakN framework. It enables end-to-end development of unique infrastructure defect detectors on digital images, achieving an accuracy of above 90%. The framework also supports the semi-automatic creation of new datasets and has modest computing power requirements. It can be used to immediately implement deep learning in the process of infrastructure management and, due to its architecture based on transfer learning, allows low metric loss when used across different datasets. We also demonstrate that thanks to its scalability and modular structure, the presented framework is easily modifiable, and can be used in realistic scenarios.&lt;/p&gt;&#13;
</style></abstract><language><style face="normal" font="default" size="100%">eng</style></language></record><record><source-app name="Bibcite" version="8.x">Drupal-Bibcite</source-app><ref-type>5</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Mateusz Żarski</style></author><author><style face="normal" font="default" size="100%">Bartosz Wójcik</style></author><author><style face="normal" font="default" size="100%">Kamil Książek</style></author><author><style face="normal" font="default" size="100%">Marek Salamak</style></author><author><style face="normal" font="default" size="100%">Jaroslaw Miszczak</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">Advances in applicable deep-learning based defect  detection</style></title></titles><keywords/><dates><year><style face="normal" font="default" size="100%">2021</style></year></dates><secondary-title><style face="normal" font="default" size="100%">2nd Workshop on Engineering  Optimization – WEO 2021</style></secondary-title><urls><style face="normal" font="default" size="100%">http://v4shm.ippt.pan.pl/WEO2021-BookOfAbstracts.pdf</style></urls><isbn><style face="normal" font="default" size="100%">987-83-65550-23-1</style></isbn><publisher><style face="normal" font="default" size="100%">Institute of Fundamental Technological Research of the Polish Academy of Sciences</style></publisher><pub-location><style face="normal" font="default" size="100%">Warsaw</style></pub-location><language><style face="normal" font="default" size="100%">eng</style></language></record><record><source-app name="Bibcite" version="8.x">Drupal-Bibcite</source-app><ref-type>5</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Bartosz Wójcik</style></author><author><style face="normal" font="default" size="100%">Mateusz Żarski</style></author><author><style face="normal" font="default" size="100%">Marek Salamak</style></author><author><style face="normal" font="default" size="100%">Jaroslaw Miszczak</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">Extracting crack characteristics from RGB-D images</style></title></titles><keywords/><dates><year><style face="normal" font="default" size="100%">2021</style></year></dates><secondary-title><style face="normal" font="default" size="100%">2nd Workshop on Engineering  Optimization – WEO 2021</style></secondary-title><urls><style face="normal" font="default" size="100%">http://v4shm.ippt.pan.pl/WEO2021-BookOfAbstracts.pdf</style></urls><isbn><style face="normal" font="default" size="100%">987-83-65550-23-1</style></isbn><publisher><style face="normal" font="default" size="100%">Institute of Fundamental Technological Research of the Polish Academy of Sciences</style></publisher><pub-location><style face="normal" font="default" size="100%">Warsaw</style></pub-location><language><style face="normal" font="default" size="100%">eng</style></language></record><record><source-app name="Bibcite" version="8.x">Drupal-Bibcite</source-app><ref-type>17</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Mateusz Żarski</style></author><author><style face="normal" font="default" size="100%">Bartosz Wójcik</style></author><author><style face="normal" font="default" size="100%">Jaroslaw Miszczak</style></author><author><style face="normal" font="default" size="100%">Bartlomiej Blachowski</style></author><author><style face="normal" font="default" size="100%">Mariusz Ostrowski</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">Computer Vision Based Inspection on Post-Earthquake With UAV Synthetic Dataset</style></title></titles><keywords/><dates><year><style face="normal" font="default" size="100%">2022</style></year><pub-dates><date><style face="normal" font="default" size="100%">10/2022</style></date></pub-dates></dates><secondary-title><style face="normal" font="default" size="100%">IEEE Access</style></secondary-title><urls><style face="normal" font="default" size="100%">https://arxiv.org/abs/2210.05282</style></urls><electronic-resource-num><style face="normal" font="default" size="100%">10.1109/ACCESS.2022.3212918</style></electronic-resource-num><volume><style face="normal" font="default" size="100%">10</style></volume><section><style face="normal" font="default" size="100%">108134</style></section><abstract><style face="normal" font="default" size="100%">&lt;div&gt;The area affected by the earthquake is vast and often difficult to entirely cover, and the earthquake itself is a sudden event that causes multiple defects simultaneously, that cannot be effectively traced using traditional, manual methods. This article presents an innovative approach to the problem of detecting damage after sudden events by using an interconnected set of deep machine learning models organized in a single pipeline and allowing for easy modification and swapping models seamlessly. Models in the pipeline were trained with a synthetic dataset and were adapted to be further evaluated and used with unmanned aerial vehicles (UAVs) in real-world conditions. Thanks to the methods presented in the article, it is possible to obtain high accuracy in detecting buildings defects, segmenting constructions into their components and estimating their technical condition based on a single drone flight.&lt;/div&gt;&#13;
</style></abstract><language><style face="normal" font="default" size="100%">eng</style></language></record><record><source-app name="Bibcite" version="8.x">Drupal-Bibcite</source-app><ref-type>17</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Bartosz Wójcik</style></author><author><style face="normal" font="default" size="100%">Mateusz Żarski</style></author><author><style face="normal" font="default" size="100%">Kamil Książek</style></author><author><style face="normal" font="default" size="100%">Jaroslaw Miszczak</style></author><author><style face="normal" font="default" size="100%">Mirosław Skibniewski</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">A deep learning method for hard-hat-wearing detection based on head center localization</style></title></titles><keywords/><dates><year><style face="normal" font="default" size="100%">2023</style></year><pub-dates><date><style face="normal" font="default" size="100%">10/2023</style></date></pub-dates></dates><secondary-title><style face="normal" font="default" size="100%">Bulletin of the Polish Academy of Sciences Technical Sciences</style></secondary-title><urls><style face="normal" font="default" size="100%">https://journals.pan.pl/dlibra/publication/147340/edition/128931/content</style></urls><electronic-resource-num><style face="normal" font="default" size="100%">10.24425/bpasts.2023.147340</style></electronic-resource-num><volume><style face="normal" font="default" size="100%">71 (EA)</style></volume><issue><style face="normal" font="default" size="100%">6</style></issue><abstract><style face="normal" font="default" size="100%">&lt;p&gt;In recent years, a lot of attention is paid to deep learning methods in the context of vision-based construction site safety systems. However, there is still more to be done to establish the relationship between supervised construction workers and their essential personal protective equipment, like hard hats. A deep learning method combining object detection, head center localization, and simple rule-based reasoning is proposed in this article. In tests, this solution surpassed the previous methods based on the relative bounding box position of different instances and direct detection of hard hat wearers and non-wearers. Achieving MS COCO style overall AP of 67.5% compared to 66.4% and 66.3% achieved by the approaches mentioned above, with class-specific AP for hard hat non-wearers of 64.1% compared to 63.0% and 60.3%. The results show that using deep learning methods with a humanly interpretable rule-based algorithm is better suited for detecting hard hat non-wearers.&lt;/p&gt;&#13;
</style></abstract><language><style face="normal" font="default" size="100%">eng</style></language></record><record><source-app name="Bibcite" version="8.x">Drupal-Bibcite</source-app><ref-type>47</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Samuel Kessler</style></author><author><style face="normal" font="default" size="100%">Mateusz Ostaszewski</style></author><author><style face="normal" font="default" size="100%">Michał Bortkiewicz</style></author><author><style face="normal" font="default" size="100%">Mateusz Żarski</style></author><author><style face="normal" font="default" size="100%">Maciej Wołczyk</style></author><author><style face="normal" font="default" size="100%">Jack Parker-Holder</style></author><author><style face="normal" font="default" size="100%">Stephen Roberts</style></author><author><style face="normal" font="default" size="100%">Piotr Miłoś</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">The Effectiveness of World Models for Continual Reinforcement Learning</style></title></titles><keywords/><dates><year><style face="normal" font="default" size="100%">2023</style></year></dates><secondary-title><style face="normal" font="default" size="100%">Conference on Lifelong Learning Agents</style></secondary-title><urls><style face="normal" font="default" size="100%">https://lifelong-ml.cc/online_proceedings</style></urls><abstract><style face="normal" font="default" size="100%">&lt;p&gt;World models power some of the most efficient reinforcement learning algorithms. In this work, we showcase that they can be harnessed for continual learning – a situation when the agent faces changing environments. World models typically employ a replay buffer for training, which can be naturally extended to continual learning. We systematically study how different selective experience replay methods affect performance, forgetting, and transfer. We also provide recommendations regarding various modeling options for using world models. The best set of choices is called Continual-Dreamer, it is task-agnostic and utilizes the world model for continual exploration. Continual-Dreamer is sample efficient and outperforms state-of-the-art task-agnostic continual reinforcement learning methods on Minigrid and Minihack benchmarks.&lt;/p&gt;&#13;
</style></abstract><pub-location><style face="normal" font="default" size="100%">Montreal, Canada</style></pub-location><language><style face="normal" font="default" size="100%">eng</style></language></record><record><source-app name="Bibcite" version="8.x">Drupal-Bibcite</source-app><ref-type>47</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Bartlomiej Blachowski</style></author><author><style face="normal" font="default" size="100%">Mariusz Ostrowski</style></author><author><style face="normal" font="default" size="100%">Mateusz Żarski</style></author><author><style face="normal" font="default" size="100%">Bartosz Wójcik</style></author><author><style face="normal" font="default" size="100%">Piotr Tauzowski</style></author><author><style face="normal" font="default" size="100%">Lukasz Jankowski</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">An Efficient Computer Vision-Based Method for Estimation of Dynamic Displacements in Spatial Truss Structures</style></title></titles><keywords/><dates><year><style face="normal" font="default" size="100%">2022</style></year><pub-dates><date><style face="normal" font="default" size="100%">06/2022</style></date></pub-dates></dates><secondary-title><style face="normal" font="default" size="100%">Workshop on Structural Health Monitoring</style></secondary-title><electronic-resource-num><style face="normal" font="default" size="100%">https://doi.org/10.1007/978-3-031-07258-1_49</style></electronic-resource-num><abstract><style face="normal" font="default" size="100%">&lt;p&gt;In the present study a comparison of frequently used computer vision (CV)-based methods for structural health monitoring of truss structures is shown. The attention is paid to template matching methods that can be classified into one of two groups: area-based and feature-based methods. Synthetic but realistic video is used in this study. Results of the comparison are reliable due to the fact that the exact displacements are known from the finite element model of the investigated structure. From the variety of tested CV methods, the Kanade–Lucas–Tomasi algorithm with FREAK-based repetitive correction outperforms the remaining tested methods in terms of the computation time with a negligibly greater estimation error.&lt;/p&gt;&#13;
</style></abstract><isbn><style face="normal" font="default" size="100%">978-3-031-07257-4</style></isbn><publisher><style face="normal" font="default" size="100%">Springer Link</style></publisher><language><style face="normal" font="default" size="100%">eng</style></language></record><record><source-app name="Bibcite" version="8.x">Drupal-Bibcite</source-app><ref-type>17</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Dariusz Pojda</style></author><author><style face="normal" font="default" size="100%">Mateusz Żarski</style></author><author><style face="normal" font="default" size="100%">Agnieszka Anna Tomaka</style></author><author><style face="normal" font="default" size="100%">Leszek Luchowski</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">dpVision: environment for multimodal images</style></title></titles><keywords/><dates><year><style face="normal" font="default" size="100%">2025</style></year><pub-dates><date><style face="normal" font="default" size="100%">02/2025</style></date></pub-dates></dates><secondary-title><style face="normal" font="default" size="100%">SoftwareX</style></secondary-title><urls><style face="normal" font="default" size="100%">http://dx.doi.org/10.1016/j.softx.2025.102093</style></urls><electronic-resource-num><style face="normal" font="default" size="100%">10.1016/j.softx.2025.102093</style></electronic-resource-num><volume><style face="normal" font="default" size="100%">30</style></volume><issue><style face="normal" font="default" size="100%">102093</style></issue><language><style face="normal" font="default" size="100%">eng</style></language></record><record><source-app name="Bibcite" version="8.x">Drupal-Bibcite</source-app><ref-type>17</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Mateusz Żarski</style></author><author><style face="normal" font="default" size="100%">Jaroslaw Miszczak</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">Multi-Step Feature Fusion for Natural Disaster Damage Assessment on Satellite Images</style></title></titles><keywords/><dates><year><style face="normal" font="default" size="100%">2024</style></year><pub-dates><date><style face="normal" font="default" size="100%">09/2024</style></date></pub-dates></dates><secondary-title><style face="normal" font="default" size="100%">IEEE Access</style></secondary-title><electronic-resource-num><style face="normal" font="default" size="100%">10.1109/ACCESS.2024.3459424</style></electronic-resource-num><section><style face="normal" font="default" size="100%">140072 - 140081</style></section><abstract><style face="normal" font="default" size="100%">Quick and accurate assessment of the damage state of buildings after natural disasters is crucial for undertaking properly targeted rescue and subsequent recovery operations, which can have a major impact on the safety of victims and the cost of disaster recovery. The quality of such a process can be significantly improved by harnessing the potential of machine learning methods in computer vision. This paper presents a novel damage assessment method using an original multi-step feature fusion network for the classification of the damage state of buildings based on pre- and post-disaster large-scale satellite images. We introduce a novel convolutional neural network (CNN) module that performs feature fusion at multiple network levels between pre- and post-disaster images in the horizontal and vertical directions of CNN network. An additional network element – Fuse Module – was proposed to adapt any CNN model to analyze image pairs in the issue of pair classification. We use, open, large-scale datasets (IDA-BD and xView2) to verify, that the proposed method is suitable to improve on existing state-of-the-art architectures. We report over a 3 percentage point increase in the accuracy of the Vision Transformer model.</style></abstract><language><style face="normal" font="default" size="100%">eng</style></language></record><record><source-app name="Bibcite" version="8.x">Drupal-Bibcite</source-app><ref-type>17</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Katarzyna Filus</style></author><author><style face="normal" font="default" size="100%">Michał Romaszewski</style></author><author><style face="normal" font="default" size="100%">Mateusz Żarski</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">Semantic Depth Matters: Explaining Errors of Deep Vision Networks through Perceived Class Similarities</style></title></titles><keywords/><dates/><secondary-title><style face="normal" font="default" size="100%">arXiv preprint</style></secondary-title><urls><style face="normal" font="default" size="100%">arXiv:2504.09956</style></urls><electronic-resource-num><style face="normal" font="default" size="100%"> https://doi.org/10.48550/arXiv.2504.09956</style></electronic-resource-num><language><style face="normal" font="default" size="100%">eng</style></language></record><record><source-app name="Bibcite" version="8.x">Drupal-Bibcite</source-app><ref-type>47</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Mateusz Żarski</style></author><author><style face="normal" font="default" size="100%">Sławomir Nowaczyk</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">Balancing Performance and Scalability of Demand Forecasting ML Models</style></title></titles><keywords/><dates><year><style face="normal" font="default" size="100%">2025</style></year><pub-dates><date><style face="normal" font="default" size="100%">05/2025</style></date></pub-dates></dates><secondary-title><style face="normal" font="default" size="100%">Intelligent Data Analysis XXIII [IDA]</style></secondary-title><urls><style face="normal" font="default" size="100%">https://doi.org/10.1007/978-3-031-91398-3_10</style></urls><electronic-resource-num><style face="normal" font="default" size="100%">10.1007/978-3-031-91398-3_10</style></electronic-resource-num><abstract><style face="normal" font="default" size="100%">&lt;p&gt;Balancing performance and scalability is a major concern when developing robust ML models for diverse, big-data scenarios, such as predicting demand for a number of products across multiple locations. The two mutually opposite approaches are to use a single ML model for maximizing scalability, often at the expense of performance, or to use a specialized model for each specific use case, which is often prohibitive in terms of computational costs. In this paper, we propose to balance those two approaches using our methods of model clustering and grouping. We achieve the performance level of a single use-case model while preserving the global scalability of the solution. In our experiments, we use a publicly available demand forecasting dataset as a use case. We develop and train baseline shallow ML models and DL models for both maximizing performance and scalability. Then, we showcase a desirable balance that can be achieved using our proposed methods, one that outperforms both shallow ML and specific use-case models.&lt;/p&gt;&#13;
</style></abstract><isbn><style face="normal" font="default" size="100%">978-3-031-91398-3</style></isbn><publisher><style face="normal" font="default" size="100%">Springer Nature</style></publisher><pub-location><style face="normal" font="default" size="100%">Konstanz, Germany</style></pub-location><language><style face="normal" font="default" size="100%">eng</style></language></record><record><source-app name="Bibcite" version="8.x">Drupal-Bibcite</source-app><ref-type>17</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Katarzyna Filus</style></author><author><style face="normal" font="default" size="100%">Mateusz Żarski</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">Beyond Accuracy: Uncovering the Role of Similarity Perception and its Alignment with Semantics in Supervised Learning</style></title></titles><keywords/><dates/><secondary-title><style face="normal" font="default" size="100%">arXiv preprint</style></secondary-title><urls><style face="normal" font="default" size="100%">https://arxiv.org/abs/2505.21338</style></urls><electronic-resource-num><style face="normal" font="default" size="100%">https://arxiv.org/abs/2505.21338</style></electronic-resource-num><abstract><style face="normal" font="default" size="100%">&lt;p&gt;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.&lt;/p&gt;&#13;
</style></abstract><language><style face="normal" font="default" size="100%">eng</style></language></record><record><source-app name="Bibcite" version="8.x">Drupal-Bibcite</source-app><ref-type>17</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Mateusz Żarski</style></author><author><style face="normal" font="default" size="100%">Sławomir Nowaczyk</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">Neuroplasticity-inspired dynamic ANNs for multi-task demand forecasting</style></title></titles><keywords/><dates><year><style face="normal" font="default" size="100%">2026</style></year><pub-dates><date><style face="normal" font="default" size="100%">07/2026</style></date></pub-dates></dates><secondary-title><style face="normal" font="default" size="100%">IEEE Access</style></secondary-title><urls><style face="normal" font="default" size="100%">https://ieeexplore.ieee.org/abstract/document/11592373</style></urls><electronic-resource-num><style face="normal" font="default" size="100%">10.1109/ACCESS.2026.3709101</style></electronic-resource-num><abstract><style face="normal" font="default" size="100%">&lt;p&gt;This paper introduces a novel approach to Dynamic Artificial Neural Networks (D-ANNs) for multi-task demand forecasting called Neuroplastic Multi-Task Network (NMT-Net). Unlike conventional methods focusing on inference-time dynamics or computational efficiency, our proposed method enables structural adaptability of the computational graph during training, inspired by neuroplasticity as seen in biological systems. Each new task triggers a dynamic network adaptation, including similarity-based task identification and selective training of candidate ANN heads, which are then assessed and integrated into the model based on their performance. We evaluated our framework using three real-world multi-task demand forecasting datasets from Kaggle. We demonstrated its superior performance and consistency, achieving lower RMSE and standard deviation compared to traditional baselines and state-of-the-art multi-task learning methods. NMT-Net offers a scalable, adaptable solution for multi-task and continual learning in time series prediction. The complete code for NMT-Net is available from our GitHub repository.&lt;/p&gt;&#13;
</style></abstract><language><style face="normal" font="default" size="100%">eng</style></language></record><record><source-app name="Bibcite" version="8.x">Drupal-Bibcite</source-app><ref-type>17</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Mariusz Ostrowski</style></author><author><style face="normal" font="default" size="100%">Bartlomiej Blachowski</style></author><author><style face="normal" font="default" size="100%">Bartosz Wójcik</style></author><author><style face="normal" font="default" size="100%">Mateusz Żarski</style></author><author><style face="normal" font="default" size="100%">Piotr Tauzowski</style></author><author><style face="normal" font="default" size="100%">Łukasz Jankowski</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">A framework for computer vision-based health monitoring of a truss structure subjected to unknown excitations</style></title></titles><keywords/><dates><year><style face="normal" font="default" size="100%">2023</style></year><pub-dates><date><style face="normal" font="default" size="100%">01/2023</style></date></pub-dates></dates><secondary-title><style face="normal" font="default" size="100%">Earthquake Engineering and Engineering Vibration</style></secondary-title><urls><style face="normal" font="default" size="100%">https://link.springer.com/article/10.1007/s11803-023-2154-3</style></urls><electronic-resource-num><style face="normal" font="default" size="100%">10.1007/s11803-023-2154-3</style></electronic-resource-num><abstract><style face="normal" font="default" size="100%">&lt;p&gt;Computer vision (CV) methods for measurement of structural vibration are less expensive, and their application is more straightforward than methods based on sensors that measure physical quantities at particular points of a structure. However, CV methods produce significantly more measurement errors. Thus, computer vision-based structural health monitoring (CVSHM) requires appropriate methods of damage assessment that are robust with respect to highly contaminated measurement data. In this paper a complete CVSHM framework is proposed, and three damage assessment methods are tested. The first is the augmented inverse estimate (AIE), proposed by Peng et al. in 2021. This method is designed to work with highly contaminated measurement data, but it fails with a large noise provided by CV measurement. The second method, as proposed in this paper, is based on the AIE, but it introduces a weighting matrix that enhances the conditioning of the problem. The third method, also proposed in this paper, introduces additional constraints in the optimization process; these constraints ensure that the stiff ness of structural elements can only decrease. Both proposed methods perform better than the original AIE. The latter of the two proposed methods gives the best results, and it is robust with respect to the selected coefficients, as required by the algorithm.&lt;/p&gt;&#13;
</style></abstract><language><style face="normal" font="default" size="100%">eng</style></language></record><record><source-app name="Bibcite" version="8.x">Drupal-Bibcite</source-app><ref-type>13</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Anna Zawadzka</style></author><author><style face="normal" font="default" size="100%">Przemysław Głomb</style></author><author><style face="normal" font="default" size="100%">Mateusz Żarski</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">Development of an effective method for detecting small buildings in satellite images</style></title></titles><keywords/><dates><year><style face="normal" font="default" size="100%">2025</style></year></dates><secondary-title><style face="normal" font="default" size="100%">PP-RAI 2025</style></secondary-title><abstract><style face="normal" font="default" size="100%">&lt;p&gt;Detecting and monitoring small-scale buildings is a critical challenge in various sectors and can be crucial for various important tasks. Buildings considered small are those with a footprint smaller than 10 × 10 meters, a definition justified by the resolution of Sentinel-2 images, which is 10 meters at best. Examples of tasks for which small building detection is essential include assessment and monitoring of damage caused by natural disasters or war, identification of illegal settlements, detecting new buildings in key military areas, and identification of structures in protected areas. Regrettably, widely available satellite imagery, such as Sentinel-2, provides very limited spatial resolution, making it impossible to efficiently detect small structures. Despite its importance, this issue remains unexplored, with very limited research directly concerning building segmentation and detecting small structures. Traditional segmentation and classification methods exhibit severe limitations, particularly due to small buildings frequently being positioned at pixel boundaries, essentially occupying an insignificant fraction of the original pixel. Additionally, in the context of buildings in complex environments (e.g., surrounded by vegetation or shadows), popular methods such as U-Net with ResNet often fail to deliver reliable results. In this study, we propose a novel approach to this issue by integrating a state-of-the-art object detection model with a deep-learning-based super-resolution network.&lt;/p&gt;&#13;
</style></abstract><language><style face="normal" font="default" size="100%">eng</style></language></record><record><source-app name="Bibcite" version="8.x">Drupal-Bibcite</source-app><ref-type>17</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Anna Zawadzka</style></author><author><style face="normal" font="default" size="100%">Mateusz Żarski</style></author><author><style face="normal" font="default" size="100%">Kamil Drejer</style></author><author><style face="normal" font="default" size="100%">Przemysław Głomb</style></author><author><style face="normal" font="default" size="100%">Michał Romaszewski</style></author><author><style face="normal" font="default" size="100%">Michał Cholewa</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%"> Map-Guided Cross-Training for Building Detection</style></title></titles><keywords/><dates><year><style face="normal" font="default" size="100%">2026</style></year></dates><secondary-title><style face="normal" font="default" size="100%">Geoscience and Remote Sensing Letters</style></secondary-title><language><style face="normal" font="default" size="100%">eng</style></language></record><record><source-app name="Bibcite" version="8.x">Drupal-Bibcite</source-app><ref-type>47</ref-type><contributors><authors><author><style face="normal" font="default" size="100%">Katarzyna Filus</style></author><author><style face="normal" font="default" size="100%">Mateusz Żarski</style></author></authors></contributors><titles><title><style face="normal" font="default" size="100%">Inspecting Training Dynamics of Similarity Development in Supervised Vision Networks </style></title></titles><keywords/><dates><year><style face="normal" font="default" size="100%">2026</style></year></dates><secondary-title><style face="normal" font="default" size="100%">European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases</style></secondary-title></record></records></xml>
