TY - JOUR AU - Harangi, Balázs AU - Baran, Ágnes AU - Beregi-Kovács, Marcell AU - Hajdu, András TI - Composing Diverse Ensembles of Convolutional Neural Networks by Penalization JF - MATHEMATICS J2 - MATHEMATICS-BASEL VL - 11 PY - 2023 IS - 23 SP - 4730 SN - 2227-7390 DO - 10.3390/math11234730 UR - https://m2.mtmt.hu/api/publication/34395482 ID - 34395482 AB - Ensemble-based systems are well known to have the capacity to outperform individual approaches if the ensemble members are sufficiently accurate and diverse. This paper investigates how an efficient ensemble of deep convolutional neural networks (CNNs) can be created by forcing them to adjust their parameters during the training process to increase diversity in their decisions. As a new theoretical approach to reach this aim, we join the member neural architectures via a fully connected layer and insert a new correlation penalty term in the loss function to obstruct their similar operation. With this complementary term, we implement the standard guideline of ensemble creation to increase the members’ diversity for CNNs in a more detailed and flexible way than similar existing techniques. As for applicability, we show that our approach can be efficiently used in various classification tasks. More specifically, we demonstrate its performance in challenging medical image analysis and natural image classification problems. Besides the theoretical considerations and foundations, our experimental findings suggest that the proposed technique is competitive. Namely, on the one hand, the classification rate of the ensemble trained in this way outperformed all the individual accuracies of the state-of-the-art member CNNs according to the standard error functions of these application domains. On the other hand, it is also validated that the ensemble members get more diverse and their accuracies are raised by adding the penalization term. Moreover, we performed a full comparative analysis, including other state-of-the-art ensemble-based approaches recommended for the same classification tasks. This comparative study also confirmed the superiority of our method, as it overcame the current solutions. LA - English DB - MTMT ER - TY - JOUR AU - Hoffmann, Miklós AU - Juhász, Imre AU - Troll, Ede Mátyás TI - Caustics of developable surfaces JF - FRONTIERS OF INFORMATION TECHNOLOGY & ELECTRONIC ENGINEERING J2 - FRONT INFORM TECH EL VL - 23 PY - 2022 IS - 3 SP - 479 EP - 487 PG - 9 SN - 2095-9184 DO - 10.1631/FITEE.2000613 UR - https://m2.mtmt.hu/api/publication/31989219 ID - 31989219 N1 - Cited By :1 Export Date: 24 March 2023 Correspondence Address: Hoffmann, M.; Institute of Mathematics and Computer Science, Hungary; email: hoffmann.miklos@uni-eszterhazy.hu Funding details: European Commission, EC Funding details: European Social Fund, ESF, EFOP-3.6.3-VEKOP-16-2017-00002 Funding text 1: Project supported by the European Union and the European Social Fund (No. EFOP-3.6.3-VEKOP-16-2017-00002). Open Access funding provided by European Union and the European Social Fund AB - While considering a mirror and light rays coming either from a point source or from infinity, the reflected light rays may have an envelope, called a caustic curve. In this paper, we study developable surfaces as mirrors. These caustic surfaces, described in a closed form, are also developable surfaces of the same type as the original mirror surface. We provide efficient, algorithmic computation to find the caustic surface of each of the three types of developable surfaces (cone, cylinder, and tangent surface of a spatial curve). We also provide a potential application of the results in contemporary free-form architecture design. LA - English DB - MTMT ER - TY - CHAP AU - Mester, Tamás AU - Szabó, György Emőd AU - Balla, Dániel ED - Karátson, Dávid ED - Nagy, Balázs TI - A talajvíz tisztulási folyamatainak értékelése a szennyvízcsatorna-hálózat kiépítését követő időszakban többváltozós statisztikai elemzések és GIS alkalmazásával T2 - X. Magyar Földrajzi Konferencia = 10th Hungarian Geographical Conference : absztraktkötet PB - Földgömb az Expedíciós Kutatásért Alapítvány CY - Budapest SN - 9786150128993 PY - 2021 SP - 1 EP - 1 PG - 1 UR - https://m2.mtmt.hu/api/publication/32287317 ID - 32287317 LA - Hungarian DB - MTMT ER - TY - JOUR AU - Gergő, Karancsi AU - Kiss, Emőke AU - Dániel, Béres AU - Balla, Dániel TI - Case study for estimation of the amount of contaminants stored in soil in an industrial area JF - ACTA GEOGRAPHICA DEBRECINA LANDSCAPE AND ENVIRONMENT SERIES J2 - ACTA GEOGR DEBRECINA LANDSCAPE VL - 15 PY - 2021 IS - 2 SP - 1 EP - 11 PG - 11 SN - 1789-4921 DO - 10.21120/LE/15/2/1 UR - https://m2.mtmt.hu/api/publication/32152091 ID - 32152091 LA - English DB - MTMT ER - TY - JOUR AU - Mester, Tamás AU - Szabó, György Emőd AU - Balla, Dániel TI - Assessment of Shallow Groundwater Purification Processes after the Construction of a Municipal Sewerage Network JF - WATER J2 - WATER-SUI VL - 13 PY - 2021 IS - 14 SN - 2073-4441 DO - 10.3390/w13141946 UR - https://m2.mtmt.hu/api/publication/32104194 ID - 32104194 LA - English DB - MTMT ER - TY - CHAP AU - Tornai, Róbert AU - Fürjes-Benke, Péter ED - Fazekas, István ED - Hajdu, András ED - Tómács, Tibor TI - Compute Shader in Image Processing Development T2 - Proceedings of the 1st Conference on Information Technology and Data Science CITDS 2020) PB - CEUR Workshop Proceedings CY - Debrecen T3 - CEUR Workshop Proceedings, ISSN 1613-0073 ; 2874. PY - 2021 SP - 218 EP - 225 PG - 8 UR - https://m2.mtmt.hu/api/publication/32073422 ID - 32073422 AB - This paper will present the OpenGL compute shader implementation of the BlackRoom software. BlackRoom is a platform-independent image processing program, which supports multiple execution branches like Vulkan fragment shader, OpenGL fragment shader, and CPU-based rendering. In order to support a wider range of devices with different amounts of memory, users can utilize tile rendering, and the program can be run in browsers thanks to the WebAssembly format. Thanks to our program’s built-in benchmark system, the performance differences between the implemented CPU- and GPU-based executing branches can be easily determined. We made a comprehensive comparison between the rendering performance of our CPU, OpenGL compute shader, fragment shader and Vulkan fragment shader branches. Latter is under development, which induces a relatively higher runtime presently. A further aim is to optimize our algorithms, which are using Vulkan API. Besides that, the program will be capable of rendering multiple effects at once with Vulkan fragment shader. Furthermore, the available GPU rendering and multi-threading features are planned to be enabled for WebAssembly platform yielded by the Qt framework [2]. LA - English DB - MTMT ER - TY - JOUR AU - Barkat, Ayoub AU - Bouaicha, Foued AU - Bouteraa, Oualid AU - Mester, Tamás AU - Ata, Behnam AU - Balla, Dániel AU - Rahal, Zakaria AU - Szabó, György Emőd TI - Assessment of Complex Terminal Groundwater Aquifer for Different Use of Oued Souf Valley (Algeria) Using Multivariate Statistical Methods, Geostatistical Modeling, and Water Quality Index JF - WATER J2 - WATER-SUI VL - 13 PY - 2021 IS - 11 PG - 26 SN - 2073-4441 DO - 10.3390/w13111609 UR - https://m2.mtmt.hu/api/publication/32062352 ID - 32062352 AB - This research aims to assess the hydrogeochemical evolution of the groundwater in Oued souf valley for drinking and irrigation purposes. To achieve this, 49 groundwater samples from the complex terminal were examined and treated concurrently with multivariate statistical methods, geostatistical modeling and the WQI (water quality index). Focusing on the physico-chemical parameters, Q mode clustering analysis detected four major water groups, where the mineralization augmented from group 1 to group 4. The hydro-chemical type was the same, Ca-Mg-Cl-SO4 for all the groups. Calcite, dolomite, anhydrite, and gypsum would be the dominant reactions with the undersaturation of evaporates minerals, based on geochemical modeling, while the carbonate minerals are precipitating. Geostatistical analysis using ordinary Kriging demonstrated the exponential semi-variogram model fitted for EC (electrical conductivity), Ca2+ (calcium), Mg2+ (magnesium), K+ (potassium), HCO3− (bicarbonate), Cl− (chloride), and SO42− (sulfate). At the same time, the rational quadratic model was the best-fitted semi-variogram model for Na+ (sodium) and NO3− (nitrate). EC, SO42−, and NO3− have a strong spatial structure, while Ca2+, Na+, K+, and HCO3− have a moderate spatial structure. Moreover, there was a weak spatial structure for Mg2+ and Cl−. The WQI shows that CT (complex terminal groundwater aquifers) are not suitable for drinking and their quality for irrigation fluctuates from excellent to moderate quality. LA - English DB - MTMT ER - TY - JOUR AU - Csoba, István AU - Kunkli, Roland Imre TI - Efficient Rendering of Ocular Wavefront Aberrations using Tiled Point‐Spread Function Splatting JF - COMPUTER GRAPHICS FORUM J2 - COMPUT GRAPH FORUM VL - 40 PY - 2021 IS - 6 SP - 182 EP - 199 PG - 18 SN - 0167-7055 DO - 10.1111/cgf.14267 UR - https://m2.mtmt.hu/api/publication/32035477 ID - 32035477 LA - English DB - MTMT ER - TY - CHAP AU - Tóth, Róbert AU - Zichar, Marianna AU - Hoffmann, Miklós ED - Cheng, Liang-Yee TI - Improving and Measuring Spatial Skills with Augmented Reality and Gamification T2 - ICGG 2020 - Proceedings of the 19th International Conference on Geometry and Graphics PB - Springer Netherlands CY - Berlin SN - 9783030634025 T3 - Advances in Intelligent Systems and Computing, ISSN 2194-5357 ; 1296. PY - 2021 SP - 755 EP - 764 PG - 10 DO - 10.1007/978-3-030-63403-2_68 UR - https://m2.mtmt.hu/api/publication/31687556 ID - 31687556 N1 - Conference code: 252709 Cited By :4 Export Date: 6 December 2022 Correspondence Address: Tóth, R.; Faculty of Informatics, Hungary; email: toth.robert@inf.unideb.hu Funding details: European Commission, EC Funding details: European Social Fund, ESF Funding text 1: Acknowledgement. This work was supported by the construction EFOP-3.6.3-VEKOP-16-2017-00002. The project was supported by the European Union, co-financed by the European Social Fund. To provide our charts and diagrams we use the AnyChart Android Chart data visualization library with educational license. LA - English DB - MTMT ER - TY - CONF AU - Erős, Bence Dániel AU - Kunkli, Roland Imre TI - A WebGL-based virtual puzzle game for spatial skill development purposes T2 - The 1st Conference on Information Technology and Data Science PY - 2020 SP - 56 EP - 58 PG - 3 UR - https://m2.mtmt.hu/api/publication/32066828 ID - 32066828 LA - English DB - MTMT ER -