Wyniki wyszukiwania - Biblioteka Nauki

1

Deep learning classification and recognition method for milling surface roughness combined with simulation data

100%

Lu L. , Yi H. , Shu A. , Qin J. , Lu E.

|

2023

|

tom Vol. 30, nr 1

117--138

EN

To address the problem that a deep neural network needs a sufficient number of training samples to have a good prediction performance, this paper firstly used the Z-Map algorithm to generate a simulated profile of the milling surface and construct an optical simulation model of surface imaging to supplement the training sample size of the neural network. Then the Deep CORAL model was used to match the textures of the simulated samples and the actual samples across domains to solve the problem that the simulated samples were not in the same domain as the actual milling samples. Experimental results have shown that high texture matching could be achieved between optical simulation images and actual images, laying the foundation for expanding the actual milled workpiece images with the simulation images. The deep convolutional neural model Xception was used to predict the classification of six classes of data sets with the inclusion of simulation images, and the accuracy was improved from 86.48% to 92.79% compared with the model without the inclusion of simulation images. The proposed method solves the problem of the need for a large number of samples for deep neural networks and lays the foundation for similar methods to predict surface roughness for different machining processes.

2

Position Estimation in Mixed Indoor-Outdoor Environment Using Signals of Opportunity and Deep Learning Approach

100%

Urwan S. , Wysocka D. R. , Pietrzak A. , Cwalina K. K.

|

tom Vol. 68. No. 3

594-607

EN

To improve the user’s localization estimation in indoor and outdoor environment a novel radiolocalization system using deep learning dedicated to work both in indoor and outdoor environment is proposed. It is based on the radio signatures using radio signals of opportunity from LTE an WiFi networks. The measurements of channel state estimators from LTE network and from WiFi network are taken by using the developed application. The user’s position is calculated with a trained neural network system’s models. Additionally the influence of various number of measurements from LTE and WiFi networks in the input vector on the positioning accuracy was examined. From the results it can be seen that using hybrid deep learning algorithm with a radio signatures method can result in localization error 24.3 m and 1.9 m lower comparing respectively to the GPS system and standalone deep learning algorithm with a radio signatures method in indoor environment. What is more, the combination of LTE and WiFi signals measurement in an input vector results in better indoor and outdoor as well as floor classification accuracy and less positioning error comparing to the input vector consisting measurements from only LTE network or from only WiFi network.

3

Track finding with Deep Neural Networks

100%

Kucharczyk M. , Wolter M.

Computer Science

|

2019

|

tom Vol. 20 (4)

475--491

EN

High energy physics experiments require fast and efficient methods for reconstructing the tracks of charged particles. The commonly used algorithms are sequential and the required CPU power increases rapidly with the number of tracks. Neural networks can speed up the process due to their capability of modeling complex non-linear data dependencies and finding all tracks in parallel. In this paper, we describe the application of the deep neural network for reconstructing straight tracks in a toy two-dimensional model. It is planned to apply this method to the experimental data obtained by the MUonE experiment at CERN.

4

End-To-End deep neural models for Automatic Speech Recognition for Polish Language

100%

Pondel-Sycz K. , Pietrzak A. P. , Szymla J.

|

2024

|

tom Vol. 70, No. 2

315--321

EN

This article concerns research on deep learning models (DNN) used for automatic speech recognition (ASR). In such systems, recognition is based on Mel Frequency Cepstral Coefficients (MFCC) acoustic features and spectrograms. The latest ASR technologies are based on convolutional neural networks (CNNs), recurrent neural networks (RNNs) and Transformers. The article presents an analysis of modern artificial intelligence algorithms adapted for automatic recognition of the Polish language. The differences between conventional architectures and ASR DNN End-To-End (E2E) models are discussed. Preliminary tests of five selected models (QuartzNet, FastConformer, Wav2Vec 2.0 XLSR, Whisper and ESPnet Model Zoo) on Mozilla Common Voice, Multilingual LibriSpeech and VoxPopuli databases are demonstrated. Tests were conducted for clean audio signal, signal with bandwidth limitation and degraded. The tested models were evaluated on the basis of Word Error Rate (WER).

5

Cooling fan controlled by embedded vision system

100%

Kula S.

|

tom No. 104

7--16

EN

The HMI (human machine interaction) systems are widely used to control machines and variety of devices. Currently the HMI solutions, based on touch screens are almost commonly used in many domains, however the number of devices, which interaction with the user is based on speech recognition or user gesture recognition increases systematically. The paper focuses on the electromechanical system, which applies gestures and handwritten digits to control the speed of the DC cooling fan. The system crucial elements are the AVR microcontroller and the developer board, equipped with the embedded supercomputer NVIDIA Jetson TX1. To create the software part of the system artificial intelligence algorithms and deep neural networks were applied. The paper describes the complete routine of data preprocessing, deep neural network training and testing with the use of the GPU Tesla K20 and with the use of the DIGITS (Deep Learning GPU Training System), deployment of the trained model on Jetson TX1 board and the system execution. The system enables to control the fan through the two gestures (“stone”, ”paper”) or through four handwritten digits.

6

Graph-based segmentation with homogeneous hue and texture vertices

88%

Ngo L. , Han J.-H.

|

tom Vol. 51, nr 4

541--549

EN

This work presents an automated segmentation method, based on graph theory, which processes superpixels that exhibit spatially similarities in hue and texture pixel groups, rather than individual pixels. The graph shortest path includes a chain of neighboring superpixels which have minimal intensity changes. This method reduces graphics computational complexity because it provides large decreases in the number of vertices as the superpixel size increases. For the starting vertex prediction, the boundary pixel in first column which is included in this starting vertex is predicted by a trained deep neural network formulated as a regression task. By formulating the problem as a regression scheme, the computational burden is decreased in comparison with classifying each pixel in the entire image. This feasibility approach, when applied as a preliminary study in electron microscopy and optical coherence tomography images, demonstrated high measures of accuracy: 0.9670 for the electron microscopy image and 0.9930 for vitreous/nerve-fiber and inner-segment/outer-segment layer segmentations in the optical coherence tomography image.

7

Power-Ground Plane Impedance Modeling Using Deep Neural Networks and an Adaptive Sampling Process

88%

Goay C. H. , Cheong Z. Q. , Low C. E. , Ahmad N. S. , Goh P.

|

tom Vol. 68. No. 4

793--798

EN

This paper proposes a deep neural network (DNN) based method for the purpose of power-ground plane impedance modeling. A composite DNN model, which is a combination of two DNNs is used to predict the Z-parameters of power ground planes from their design parameters. The first DNN predicts the normalized Z-parameters whereas the second DNN predicts the original maximum and minimum values of the nonnormalized Z-parameters. This allows the method to retain a high accuracy when predicting responses that have large variations across designs, as is the case with the Z-parameters of the power-ground planes. We use the adaptive sampling algorithm to generate the training and validation samples for the DNNs. The adaptive sampling algorithm starts with only a few samples, then slowly generates more samples in the non-linear regions within the design parameters space. The level of non-linearity of the regions is determined by a surrogate model which is also trained using the generated samples as well. If the surrogate model has poor prediction accuracy in a region, then the adaptive sampling algorithm will generate more samples in that region. A shallow neural network is used as the surrogate model for non-linearity determination of the regions since it is faster to train and update. Once all the samples have been generated, they will be used to train and validate the composite DNN models. Finally, we present two examples, a square-shaped power ground plane and a squareshaped power ground plane with a hollow square at the center to demonstrate the robustness of the DNN composite models.

8

Deep Neural Network for Supervised Single-Channel Speech Enhancement

75%

Saleem N. , Irfan Khattak M. , Ali M. Y. , Shafi M. , Saleem N. , Irfan Khattak M. , Ali M. Y. , Shafi M.

|

9

Using deep neural networksto improve the precision of fast-sampled particle timing detectors

75%

Kocot M. , Misan K. , Avati V. , Bossini E. , Grzanka L. , Minafra N.

|

tom T. 25 (1)

43--61

EN

Measurements from particle timing detectors are often affected by the timewalk effect caused by statistical fluctuations in the charge deposited by passingparticles. The constant fraction discriminator (CFD) algorithm is frequentlyused to mitigate this effect both in test setups and in running experiments,such as the CMS-PPS system at the CERN’s LHC. The CFD is simple andeffective but does not leverage all voltage samples in a time series. Its performance could be enhanced with deep neural networks, which are commonlyused for time series analysis, including computing the particle arrival time. Weevaluated various neural network architectures using data acquired at the testbeam facility in the DESY-II synchrotron, where a precise MCP (MicroChan-nel Plate) detector was installed in addition to PPS diamond timing detectors.MCP measurements were used as a reference to train the networks and com-pare the results with the standard CFD method. Ultimately, we improved thetiming precision by 8% to 23%, depending on the detector’s readout channel.The best results were obtained using a UNet-based model, which outperformedclassical convolutional networks and the multilayer perceptron.

10

A deep attention network via high-resolution representation for liver and liver tumor segmentation

75%

Li Y. , Zou B. , Liu Q.

|

2021

|

tom Vol. 41, no. 4

1518--1532

EN

The segmentation of liver and liver tumor is an essential step for computer-aided liver disease diagnosis, treatment and prognosis. Although deep convolutional neural networks have contributed to liver and tumor segmentation, their architectures can not maintain spatial details and long-range context information. Besides, the fixed receptive fields of these networks limit the segmentation performance of livers and tumors with variant sizes and shapes. To address above problems, we propose a deep attention neural network which contains high-resolution branch and multi-scale features aggregation for cascaded liver and tumor segmentation from CT images. To be specific, the high-resolution branch can maintain the resolution of the input image and thus preserves the spatial details. The multi-scale features exchange and fusion enable the receptive fields of the network to adapt to liver and tumor with variant shapes and sizes. The appended attention module evaluates the similarities between every two pixels to model the long-range dependence and context information so that the network can segment liver and tumor areas located in distant regions. Experimental results on the LiTS and the 3D-IRCADb datasets demonstrate that our method can generate satisfying performance.

11

Koncepcja architektur sieci neuronowych na potrzeby klasyfikacji sygnałów radiowych

75%

Kukliński D.

|

tom Vol. 63, Nr 6

12--17

PL

Problem automatycznej klasyfikacji widma jest coraz ważniejszy w dzisiejszych czasach. Jest to kluczowy problem, który należy rozwiązywać w systemach komunikacji bezprzewodowej wykorzystywanych w zastosowaniach wojskowych jaki i cywilnych. Istnieje wiele metod rozwiązujących powyżej przedstawione zagadnienie, tradycyjne, oparte głównie na statystyce oraz wykorzystujące algorytmy sztucznej inteligencji, które są zakwalifikowane do metod zaawansowanych. W artykule przed- stawiono koncepcje dwóch modeli głębokich sieci neuronowych, które rozwiązują problem automatycznej klasyfikacji modulacji w nowatorski sposób. Zaproponowane struktury same dostosowują przetwarzanie otrzymanego sygnału w celu detekcji rodzaju modulacji. Pierwsza struktura wykorzystuje transformatę Wignera-Vill’a, natomiast drugi model wykorzystuje krótko okresową transformatę Fouriera. Obie transformaty są zaimplementowane w warstwie wejściowej z wagami, które oddziałują na parametry tych transformat.

EN

The problem of automatic spectrum classification is becoming more and more important nowadays. This is a key problem that needs to be solved in wireless communication systems used in military and civilian applications. There are many traditional methods that solve the problem presented above, mainly based on statistics and using artificial intelligence algorithms, which are classified as advanced methods. The article presents the concepts of two models of deep neural networks that solve the problem of automatic classification of modulation in an innovative way. The proposed structures themselves adapt the processing of the received signal in order to detect the modulation type. The first structure uses the Wigner-Vill transform, while the second model uses the short period Fourier transform. Both transforms are implemented in the input layer with weights that affect their execution.

12

Wykrywanie anomalii w sieciach sterowanych programowo z wykorzystaniem dynamicznej telemetrii oraz głębokich sieci neuronowych

63%

Rzym G. , Masny A.

|

2019

|

tom nr 7

635--638, CD

PL

Wraz ze wzrostem dostępności mocy obliczeniowej uczenie maszynowe w dzisiejszych czasach skupia się coraz bardziej na metodach głębokiego uczenia. Powszechna automatyzacja procesów skłania do przemyślenia nowoczesnych implementacji sieciowych w rozumieniu ich bezpieczeństwa i szybkiego, jak i dokładnego reagowania na awarie. Niniejszy artykuł opisuje wykorzystanie głębokich sieci neuronowych do wykrywania anomalii w ruchu sieciowym w sieciach sterowanych programowo (SDN). Dodatkowo, obrazuje szerszy pogląd na automatyzację monitorowania sieci z wykorzsytaniem dynamicznej telemetrii.

EN

With the increasing availability of computational power, nowadays machine learning focuses more and more on deep learning methods. The widespread automation of processes leads to the rethinking of modern network implementations in the understanding of their safety and quick and accurate response to failures. This article describes the use of deep neural networks to detect anomalies in network traffic in conjunction with SDN. In addition, it provides a broader view of network monitoring automation with the usage of dynamic telemetry.

13

An intelligent approach to short-term wind power prediction using deep neural networks

63%

Niksa-Rynkiewicz T. , Stomma P. , Witkowska A. , Rutkowska D. , Słowik A. , Cpałka K. , Jaworek-Korjakowska J. , Kolendo P.

Journal of Artificial Intelligence and Soft Computing Research

|

2023

|

tom Vol. 13. No. 3

197--210

EN

In this paper, an intelligent approach to the Short-Term Wind Power Prediction (STWPP) problem is considered, with the use of various types of Deep Neural Networks (DNNs). The impact of the prediction time horizon length on accuracy, and the influence of temperature on prediction effectiveness have been analyzed. Three types of DNNs have been implemented and tested, including: CNN (Convolutional Neural Networks), GRU (Gated Recurrent Unit), and H-MLP (Hierarchical Multilayer Perceptron). The DNN architectures are part of the Deep Learning Prediction (DLP) framework that is applied in the Deep Learning Power Prediction System (DLPPS). The system is trained based on data that comes from a real wind farm. This is significant because the prediction results strongly depend on weather conditions in specific locations. The results obtained from the proposed system, for the real data, are presented and compared. The best result has been achieved for the GRU network. The key advantage of the system is a high effectiveness prediction using a minimal subset of parameters. The prediction of wind power in wind farms is very important as wind power capacity has shown a rapid increase, and has become a promising source of renewable energies.

14

Optimal biomarker selection using adaptive Social Ski-Driver optimization for liver cancer detection

63%

Gunasekhar P. , Vijayalakshmi S.

|

tom Vol. 40, no. 4

1611--1625

EN

Most essential biomolecule found in the human body is a biomarker; with these biomarkers, the abnormal biological processes and disease states of each patient can be accurately determined. Nowadays, the biomarker applications are frequently applied during clinical trials to identify cancer patients. In this method, the major significance of miRNA biomarkers during liver cancer detection is analysed. For such analysis, a deep learning technique is introduced along with optimization algorithms. Six different filter-based approaches are considered for feature selection they are Chi-Squared (Chi2), Information Gain (IG), Gain Ratio (GR), Symmetrical Uncertainty (SU), RelieF (RF) and RF-W. Two high ranked features from these selected features are extracted by the Modified Social Ski-Driver optimization (MSSO) algorithm. With that high ranked features, the liver cancer tissues are accurately detected by Sunflower Optimization-based deep neural network (DSFNN) approach. The analysis part concludes that a miRNA biomarker having a higher rank provide better cancer detection results than other low-ranked biomarkers. In this work, 10 different, clinically verified miRNA biomarkers are selected for this detection process. The data required for liver cancer detection is selected from NCBI-GEO database. The performance of this entire cancer detection process is evaluated by accuracy, sensitivity, precision, specificity, and Area under curve (AUC) metrics. Furthermore, we also determined that the usage of 10, 5, and 3 clinically verified miRNAs provide better cancer detection results than other miRNAs. Among all clinically verified miRNAs, the selected three biomarkers (hsa-mir-10b, hsa-let-7c, hsa-mir- 145) has attained higher recognition result. The performance result attained by the proposed DSFNN is compared with five different algorithms for both training and validation datasets.

15

Deep learning application on object tracking

63%

Taglout R. , Saoud B.

|

tom R. 99, nr 9

145--149

EN

The challenge of correctly identifying the target in the first frame of continuous sequences and tracking it in succeeding frames is frequently solved by visual tracking. The development of deep neural networks has aided in significant advancement over the past few decades. However, they are still considerable challenges in developing reliable trackers in challenging situations, essentially due to complicated backgrounds, partial or complete occlusion, illumination change, blur and similar objects. In this paper, we study correlation filter and deep learning-based approaches. We have compared the following trackers ECO, SaimRPN, ATOM, DiMP, TRASFUST and TREG. These trackers have been developed based on deep neural networks and are very recent. Performances of trackers have been evaluated on OTB-100, UAV123, VOT 2019, GOT-10k and LaSOT dataset. Results prove the effectiveness of deep neural networks to cope up with object tracking in videos.

PL

Wyzwanie polegające na prawidłowej identyfikacji celu w pierwszej klatce ciągłych sekwencji i śledzeniu go w kolejnych klatkach jest często rozwiązywane przez śledzenie wizualne. Rozwój głębokich sieci neuronowych przyczynił się do znacznego postępu w ciągu ostatnich kilku dekad. Jednak nadal stanowią one poważne wyzwanie w opracowywaniu niezawodnych trackerów w trudnych sytuacjach, głównie ze względu na skomplikowane tła, częściowe lub całkowite przesłonięcie, zmiany oświetlenia, rozmycie i podobne obiekty. W tym artykule badamy filtr korelacji i podejście oparte na głębokim uczeniu się. Porównaliśmy następujące trackery ECO, SaimRPN, ATOM, DiMP, TRASFUST i TREG. Te trackery zostały opracowane w oparciu o głębokie sieci neuronowe i są bardzo nowe. Wydajność trackerów została oceniona na zestawie danych OTB-100, UAV123, VOT 2019, GOT-10k i LaSOT. Wyniki dowodzą skuteczności głębokich sieci neuronowych w radzeniu sobie ze śledzeniem obiektów w filmach.

16

Deep Neural Networks for Curbing Climate Change-Induced Farmers-Herdsmen Clashes in a Sustainable Social Inclusion Initiative

63%

Okewu E. , Misra S. , Fernandez S. L. , Ayeni F. , Mbarika V. , Damaševičius R.

|

tom Vol. 14, nr 2

143--155

EN

Peaceful coexistence of farmers and pastoralists is becoming increasingly elusive and has adverse impact on agricultural revolution and global food security. The targets of Sustainable Development Goal 16 (SDG 16) include promoting peaceful and inclusive societies for sustainable development, providing access to justice for all and building effective, accountable and inclusive institutions at all levels. As a soft approach and long term solution to the perennial farmers-herdsmen clashes with attendant humanitarian crisis, this study proposes a social inclusion architecture using deep neural network (DNN). This is against the backdrop that formulating policies and implementing programmes based on unbiased information obtained from historical agricultural data using intelligent technology like deep neural network (DNN) can be handy in managing emotions. In this vision paper, a DNN-based Farmers-Herdsmen Expert System (FHES) is proposed based on data obtained from the Nigerian National Bureau of Statistics for tackling the incessant climate change-induced farmers-herdsmen clashes, with particular reference to Nigeria. So far, many lives have been lost. FHES is modelled as a deep neural network and trained using farmers-herdsmen historical data. Input variables used include land, water, vegetation, and implements while the output is farmers/herders disposition to peace. Regression analysis and pattern recognition performed by the DNN on the farmers-herdsmen data will enrich the inference engine of FHES with extracted rules (knowledge base). This knowledge base is then relied upon to classify future behaviours of herdsmen/farmers as well as predict their dispositions to violence. Critical stakeholders like governments, service providers and researchers can leverage on such advisory to initiate proactive and socially inclusive conflict prevention measures such as people-friendly policies, programmes and legislations. This way, conflicts can be averted, national security challenges tackled, and peaceful atmosphere guaranteed for sustainable development.

PL

Pokojowe współistnienie rolników i pasterzy staje się coraz mnie realne, co ma negatywny wpływ na rewolucję rolniczą i globalne bezpieczeństwo żywnościowe. Cele zrównoważonego rozwoju (SDG 16) obejmują promowanie tworzenia pokojowych i zintegrowanych społeczeństw na rzecz zrównoważonego rozwoju, zapewnienie wszystkim dostępu do uczciwego wymiaru sprawiedliwości i tworzenie skutecznych, odpowiedzialnych i integrujących instytucji na wszystkich poziomach. W ramach łagodnego podejścia i długofalowego podejścia do problemu konfliktów rolników-pasterzy w kontekście kryzysu humanitarnego, w niniejszym artykule zaproponowano architekturę integracji społecznej wykorzystującą głęboką sieć neuronową (DNN). Formułowanie polityki i wdrażanie programów w oparciu o obiektywne informacje uzyskane z historycznych danych przy użyciu inteligentnej technologii, takiej jak głęboka sieć neuronowa (DNN), może być przydatne w zarządzaniu emocjami. W niniejszym artykule zaproponowano oparty na danych uzyskanych od Nigeryjskiego Narodowego Urzędu Statystycznego system ekspercki rolników-pasterzy (FHES) oparty na DNN w celu przeciwdziałaniu nieustannym starciom rolników-pasterzy wywołanych zmianami klimatu, ze szczególnym uwzględnieniem Nigerii. Do tej pory wiele było ofiar. System FHES jest modelowany jako głęboka sieć neuronowa, przy użyciu danych historycznych hodowców-pasterzy. Zastosowane zmienne wejściowe obejmują ziemię, wodę, roślinność i narzędzia, podczas gdy zmienne wyjściowe to rolnicy-pasterze skłonni do pokoju. Analiza regresji i rozpoznawanie wzorców przeprowadzone przez DNN na danych rolników-pasterzy wzbogaci mechanizm wnioskowania systemu FHES o wyodrębnione reguły (baza wiedzy). Podstawą tej wiedzy jest klasyfikacja przyszłych zachowań pasterzy/rolników, a także przewidywanie ich skłonności do przemocy. Krytyczni interesariusze, tacy jak rządy, dostawcy usług i naukowcy, mogą wykorzystać takie doradztwo do zainicjowania proaktywnych i społecznie włączających środków zapobiegania konfliktom, takich jak przyjazne dla ludzi polityki, programy i prawodawstwo. W ten sposób można uniknąć konfliktów, stawić czoła wyzwaniom bezpieczeństwa narodowego i zagwarantować pokojową atmosferę dla zrównoważonego rozwoju.

17

Use of features from RR-time series and EEG signals for automated classification of sleep stages in deep neural network framework

63%

Tripathy R. K. , Rajendra Acharya U.

|

tom Vol. 38, no. 4

890--902

EN

Sleep is a physiological activity and human body restores itself from various diseases during sleep. It is necessary to get sufficient amount of sleep to have sound physiological and mental health. Nowadays, due to our present hectic lifestyle, the amount of sound sleep is reduced. It is very difficult to decipher the various stages of sleep manually. Hence, an automated systemmay be useful to detect the different stages of sleep. This paper presents a novel method for the classification of sleep stages based on RR-time series and electroencephalogram (EEG) signal. The method uses iterative filtering (IF) based multiresolution analysis approach for the decomposition of RR-time series into intrinsic mode functions (IMFs). The delta (d), theta (u), alpha (a), beta (b) and gamma (g) waves are evaluated from EEG signal using band-pass filtering. The recurrence quantification analysis (RQA) and dispersion entropy (DE) based features are evaluated from the IMFs of RR-time series. The dispersion entropy and the variance features are evaluated from the different bands of EEG signal. The RR-time series features and the EEG features coupled with the deep neural network (DNN) are used for the classification of sleep stages. The simulation results demonstrate that our proposed method has achieved an average accuracy of 85.51%, 94.03% and 95.71% for the classification of 'sleep vs wake', 'light sleep vs deep sleep' and 'rapid eye movement (REM) vs non-rapid eye movement (NREM)' sleep stages.

18

Analysis and Forecasting of the Primary Energy Consumption in Poland Using Deep Learning

63%

Manowska A.

Inżynieria Mineralna

|

2020

|

tom no. 1, vol. 1

217--222

EN

Consumption of fossil energy resources were increased dramatically, due to the economic and population growth. In turn, the consumption of fossil resources causes depletion of resources and contributes to environmental pollution. The European Union's "climate neutrality" initiative requires effective energy management from the member states. By this is meant a resource-efficient and competitive economy in which there is no greenhouse gas emission and where economic growth is decoupled from resource consumption. The article analyzes the level of primary energy consumption in Poland. It was examined whether a 23% drop in energy consumption could be achieved in 2030 compared to the base year and according with energy efficiency assumptions. A methodology for forecasting primary energy consumption based on deep neural networks, in particular on Long Short Term Memory (LSTM) algorithms was also presented.

PL

Zużycie kopalnych surowców energetycznych wzrasta, a wzrost ten jest skorelowany ze wzrostem ludności i rozwojem gospodarczym. Z kolei zużycie kopalnych surowców energetycznych powoduje wyczerpywanie się zasobów i przyczynia się do zanieczyszczenia środowiska. Inicjatywa Unii Europejskiej "neutralność klimatyczna" wymaga od państw członkowskich efektywnego zarządzania energią. Przez co rozumie się zasobooszczędną i konkurencyjną gospodarką, w której nie ma emisji netto gazów cieplarnianych i gdzie wzrost gospodarczy jest oddzielony od zużycia zasobów. W artykule przeanalizowano poziom zużycia energii pierwotnej w Polsce. Zbadano, czy w roku 2030 uda się osiągnąć 23% spadek konsumpcji energii w odniesieniu do roku bazowego, zgodnie z przyjętymi założeniami o efektywności energetycznej. Przedstawiono również metodologię prognozowania zużycia energii pierwotnej opartą na głębokich sieciach neuronowych, w szczególności na algorytmach Long Short Term Memory (LSTM).

19

Breast cancer detection from histopathology images using modified residual neural networks

51%

Gupta V. , Vasudev M. , Doegar A. , Sambyal N.

|

2021

|

tom Vol. 41, no. 4

1272--1287

EN

Breast cancer is one of the major causes of death among women worldwide. Efficient diagnosis of breast cancer in the early phases can reduce the associated morbidity and mortality and can provide a higher probability of full recovery. Computer-aided detection systems use computer technologies to detect abnormalities in clinical images which can assist medical professionals in a faster and more accurate diagnosis. In this paper, we propose a modified residual neural network-based method for breast cancer detection using histopathology images. The proposed approach provides good performance over varying magnification factors of 40X, 100X, 200X and 400X. The network obtains an average classification accuracy of 99.75%, precision of 99.18% and recall of 99.37% on BreakHis dataset with 40X magnification factor. The proposed work outperforms the existing methods and delivers state-of-the-art results on the benchmark breast cancer dataset.

20

Deep-neural-networks-based approaches for Biot-squirt model in rock physics

51%

Xiong F. , Liu J. , Guo Z. , Liu J.

Acta Geophysica

|

2022

|

tom Vol. 70, no 2

593--607

EN

A new cost-effective surrogate model using deep neural network (DNN) for seismic wave propagation in rocks saturated with fluid is presented. In this field, the dispersion/attenuation analysis and wave-field simulation are two key measurements which can be carried out by solving wave equations. The Biot–squirt (BISQ) equation is a classical wave propagation model in geophysical forward modeling and has been widely used. The solution of such equation, especially by numerical method, is often complex and time-consuming. In this work, a DNN model is trained with the dataset of velocity and inverse quality factor generated from BISQ model. The results show that the relative mean square error between the predictions of DNN model and that of BISQ model on the test sets are all less than 3%. It indicates that the DNN model has learned the high-dimensional space well and then can realize the dispersion/attenuation analysis for any given rock physical parameters. Besides, the other well-trained DNN model is used to obtain the simulation results with second-order accuracy according to results by finite difference scheme with first-order accuracy. It reveals that the fast wave-field simulation can be implemented once the results with lower accuracy are obtained.