Filtry
wszystkich: 624
wybranych: 527
-
Katalog
Filtry wybranego katalogu
Wyniki wyszukiwania dla: DEEP%20NEURAL%20NETWORKS
-
Arsenic concentrations in Baltic Sea sediments close to chemical munitions dumpsites
Publikacja -
Chemical Munitions Search & Assessment—An evaluation of the dumped munitions problem in the Baltic Sea
Publikacja -
Electrostatic interface recombination in the system of disordered materials characterized by different permittivities
PublikacjaWe report on the analysis of an electrostatic interface recombination in a system consisting of disordered organic materials. This process is a consequence of the polarization effect which takes place at the interface of two phases characterized by different permittivities. In this paper, the impact of tail and deep localized states on the recombination order is demonstrated. We also discuss the influence of temperature on this...
-
Spiral Search Grasshopper Features Selection with VGG19-ResNet50 for Remote Sensing Object Detection
PublikacjaRemote sensing object detection plays a major role in satellite imaging and is required in various scenarios such as transportation, forestry, and the ocean. Deep learning techniques provide efficient performance in remote sensing object detection. The existing techniques have the limitations of data imbalance, overfitting, and lower efficiency in detecting small objects. This research proposes the spiral search grasshopper (SSG)...
-
Stan przedawaryjny wykopu głębokiego w obudowie z pali CFA
PublikacjaW artykule przedstawiono stan przedawaryjny wykopu głębokiego w obudowie z pali CFA. Omówiono warunki geotechniczne podłoża oraz rozwiązania projektowe. Przeprowadzono analizę dwóch wersji projektów konstrukcyjnych oraz fazę wykonawstwa z wskazaniem na liczne błędy. Przedstawiono możliwe oraz przyjęte działania naprawcze. Omówiono problemy i błędy projektowe na etapie działań naprawczych. Wskazano przyczyny wystąpienia stanu przedawaryjnego.
-
Self-Supervised Learning to Increase the Performance of Skin Lesion Classification
PublikacjaTo successfully train a deep neural network, a large amount of human-labeled data is required. Unfortunately, in many areas, collecting and labeling data is a difficult and tedious task. Several ways have been developed to mitigate the problem associated with the shortage of data, the most common of which is transfer learning. However, in many cases, the use of transfer learning as the only remedy is insufficient. In this study,...
-
Application of the laser diode with central wavelength 975 nm for the therapy of neurofibroma and hemangiomas
PublikacjaThis paper presents newly developed dermatological laser (with central wavelength 975 nm) for application in therapies requiring deep penetration of tissue, e.g. cutaneous (dermal) neurofibroma (Recklinghausen disease) and hemangiomas. This laser can work either in pulse or continues wave mode. Laser radiation is transmitted toward the application region by optical fiber with a diameter of 0.6 mm. The compact design of the laser...
-
Enhanced trap-assisted recombination in organic semiconductors
PublikacjaAn analytical model to describe the interaction of excitons and charge transfer states with deep traps is formulated for the case of molecular materials. Here, we have considered the influence of a trap-assisted recombination on this phenomenon. The final expression for the effective recombination rate has been derived from the Shockley–Read–Hall theory and kinetic equations which characterize different photophysical processes....
-
SIGNIFICANT GUIDELINE FOR THE DAMAGE INDICES APPLIED TO REINFORCED CONCRETE STRUCTURES
PublikacjaIn this paper, based on a deep overview and literature study, the different formulas proposed for damage indices (DIs) applied to reinforced concrete structures under monotonic or cyclic loading are classified and presented. The DIs are applied to quantify the damages to structures, ranging from zero to one. Normally, they are applied to make a decision for repairing or demolition of the structures in the post-earthquake phases. Keywords:...
-
Comparative model tests of SDP and CFA pile groups in non-cohesive soil
PublikacjaThe research topic relates to the subject of deep foundations supported on continuous flight auger (CFA) piles and screw displacement piles (SDP). The authors have decided to conduct model tests of foundations supported on the group of piles mentioned above and also the tests of the same piles working alone. The tests are ongoing in Geotechnical Laboratory of Gdańsk University of Technology. The description of test procedure, interpretation...
-
Designing acoustic scattering elements using machine learning methods
PublikacjaIn the process of the design and correction of room acoustic properties, it is often necessary to select the appropriate type of acoustic treatment devices and make decisions regarding their size, geometry, and location of the devices inside the room under the treatment process. The goal of this doctoral dissertation is to develop and validate a mathematical model that allows predicting the effects of the application of the scattering...
-
Autonomous port management based AGV path planning and optimization via an ensemble reinforcement learning framework
PublikacjaThe rapid development of shipping trade pushes automated container terminals toward the direction of intelligence, safety and efficiency. In particular, the formulation of AGV scheduling tasks and the safety and stability of transportation path is an important part of port operation and management, and it is one of the basic tasks to build an intelligent port. Existing research mainly focuses on collaborative operation between...
-
Investigating Feature Spaces for Isolated Word Recognition
PublikacjaMuch attention is given by researchers to the speech processing task in automatic speech recognition (ASR) over the past decades. The study addresses the issue related to the investigation of the appropriateness of a two-dimensional representation of speech feature spaces for speech recognition tasks based on deep learning techniques. The approach combines Convolutional Neural Networks (CNNs) and timefrequency signal representation...
-
Active Site Architecture and Reaction Mechanism Determination of Cold Adapted beta-D-galactosidase from Arthrobacter sp. 32cB
PublikacjaArthbetaDG is a dimeric, cold-adapted beta-D-galactosidase that exhibits high hydrolytic and transglycosylation activity. A series of crystal structures of its wild form, as well as its ArthbetaDG_E441Q mutein complexes with ligands were obtained in order to describe the mode of its action. The ArthbetaDG_E441Q mutein is an inactive form of the enzyme designed to enable observation of enzyme interaction with its substrate. The...
-
Trees as a Shading System for Streets on the East–West Axis: Computer Simulations for the Selected Geometrical Proportions of Building Developments in Humid Continental Climate
PublikacjaThe study is aimed at investigating the possibilities for solar protection provided to the street canyon located on the E–W axis and with the following profiles: shallow (height/width (H/W) = 0.2, 0.6, and 1) and deep (H/W = 2) by two rows of trees located at a distance of 3 m away from southern and northern façades. The research was based on numerical simulation analyses conducted in the Rhinoceros® program, with the application...
-
Experience-Based Cognition for Driving Behavioral Fingerprint Extraction
PublikacjaABSTRACT With the rapid progress of information technologies, cars have been made increasingly intelligent. This allows cars to act as cognitive agents, i.e., to acquire knowledge and understanding of the driving habits and behavioral characteristics of drivers (i.e., driving behavioral fingerprint) through experience. Such knowledge can be then reused to facilitate the interaction between a car and its driver, and to develop better and...
-
A body position influence on ECG derived respiration
PublikacjaAn influence of a human body position on ECG derived respiration (EDR) signal is presented in the paper. Examinations were performed during deep, suspended and normal breathing for eight people in four different body positions. EDR and thoracic impedance signals were compared using correlation and standard deviation coefficients. Obtained results have shown that it is possible to monitor breath activity of people being in different...
-
Wycena środowiska w aspekcie odpowiedzialności moralnej
Publikacjathe article '' the pricing of environment in the moral responsibility context" presents ethical protection of natural environment within sustainable development paradigm as well as the problem of individual, moral responsibility in the liberal civilization. it contains different opinions on the environmental ethics, compares flat and deep ecology as well as environmental and ecological economy; it characterizes methods of economic...
-
Detection of the Oocyte Orientation for the ICSI Method Automation
PublikacjaAutomation or even computer assistance of the popular infertility treatment method: ICSI (Intracytoplasmic Sperm Injection) would speed up the whole process and improve the control of the results. This paper introduces a preliminary research for automatic spermatozoon injection into the oocyte cytoplasm. Here, the method for detection a correct orientation of the polar body of the oocyte is presented. Proposed method uses deep...
-
Investigation on the Sources and Impact of Trace Elements in the Annual Snowpack and the Firn in the Hansbreen (Southwest Spitsbergen)
PublikacjaWe present a thorough evaluation of the water soluble fraction of the trace element composition (Ca, Sr, Mg, Na, K, Li, B, Rb, U, Ni, Co, As, Cs, Cd, Mo, Se, Eu, Ba, V, Ge, Ga, Cr, Cr, P, Ti, Mn, Zr, Ce, Zn, Fe, Gd, Y, Pb, Bi, Yb, Al, Nb, Er, Nd, Dy, Sm, Ho, Th, La, Lu, Tm, Pr, Tb, Fe, In, Tl) and their fluxes in the annual snowpack and the firn of the Hansbreen (a tidewater glacier terminating in the Hornsund fjord, southwest...
-
Adversarial attack algorithm for traffic sign recognition
PublikacjaDeep learning suffers from the threat of adversarial attacks, and its defense methods have become a research hotspot. In all applications of deep learning, intelligent driving is an important and promising one, facing serious threat of adversarial attack in the meanwhile. To address the adversarial attack, this paper takes the traffic sign recognition as a typical object, for it is the core function of intelligent driving. Considering...
-
Organic solvents aggregating and shaping structural folding of protein, a case study of the protease enzyme
PublikacjaLow solubility of reactants or products in aqueous solutions can result in the enzymatic catalytic reactions that can occur in non-aqueous solutions. In current study we investigated aqueous solutions containing different organic solvents / deep eutectic solvents (DESs) that can influence the protease enzyme's activity, structural, and thermal stabilities. Retroviral aspartic protease enzyme is responsible for the cleavage of the...
-
imPlatelet classifier: image‐converted RNA biomarker profiles enable blood‐based cancer diagnostics
PublikacjaLiquid biopsies offer a minimally invasive sample collection, outperforming traditional biopsies employed for cancer evaluation. The widely used material is blood, which is the source of tumor-educated platelets. Here, we developed the imPlatelet classifier, which converts RNA-sequenced platelet data into images in which each pixel corresponds to the expression level of a certain gene. Biological knowledge from the Kyoto Encyclopedia...
-
High order of nongeminate recombination in organic bulk heterojunction solar cells
PublikacjaWe analyze high order of nongeminate recombination in organic donor–acceptor bulk heterojunction solar cells. The model of recombination where an exciton annihilates on an electron–hole Langevin bound pair near donor–acceptor interface has been applied in our studies. We obtained satisfactory agreement between experimental results and theoretical calculations for the concentration dependences of several parameters characterizing...
-
Finite element method simulations of various cases of crash tests with N2/W4/A steel road barrier
PublikacjaThe subject of this study is performance of N2/W4/A steel road safety barrier investigated in numerical simulations. System was checked under several types of initial conditions, which were assumed basing on the TB11 and TB32 normative crash tests. The main goal of present study is to investigate the relationship between initial conditions (angle and velocity) of the impact and the severity indices (associated to the vehicle occupant) during...
-
AITP - AI Thermal Pedestrians Dataset
PublikacjaEfficient pedestrian detection is a very important task in ensuring safety within road conditions, especially after sunset. One way to achieve this goal is to use thermal imaging in conjunction with deep learning methods and an annotated dataset for models training. In this work, such a dataset has been created by capturing thermal images of pedestrians in different weather and traffic conditions. All images were manually annotated...
-
A simple model of the trap-assisted recombination with the excitonic Auger mechanism
PublikacjaWe present a simple model of the trap-assisted recombination combined with the excitonic Auger mechanism. It has been shown that only six independent transitions of electrons and holes should be taken into account to describe a combination of the Shockley–Read–Hall (SRH) recombination with this excitonic process. This is in opposition to a well-known model of the SRH mechanism with the free carriers Auger effect via deep states,...
-
When Neural Networks Meet Decisional DNA: A Promising New Perspective for Knowledge Representation and Sharing
PublikacjaABSTRACT In this article, we introduce a novel concept combining neural network technology and Decisional DNA for knowledge representation and sharing. Instead of using traditional machine learning and knowledge discovery methods, this approach explores the way of knowledge extraction through deep learning processes based on a domain’s past decisional events captured by Decisional DNA. We compare our approach with kNN (k-nearest...
-
LSTM-based method for LOS/NLOS identification in an indoor environment
PublikacjaDue to the multipath propagation, harsh indoor environment significantly impacts transmitted signals which may adversely affect the quality of the radiocommunication services, with focus on the real-time ones. This negative effect may be significantly reduced (e.g. resources management and allocation) or compensated (e.g. correction of position estimation in radiolocalisation) by the LOS/NLOS identification algorithm. This paper...
-
RSS-Based DoA Estimation Using ESPAR Antenna for V2X Applications in 802.11p Frequency Band
PublikacjaIn this paper, we have proposed direction-of arrival (DoA) estimation of incoming signals for V2X applications in 802. 11p frequency band, based on recording of received signal strength (RSS) at electronically steerable parasitic array radiator (ESPAR) antenna's output port. The motivation of the work was to prove that ESPAR antenna used to increase connectivity and security in V2X communication can be also used for DoA estimation....
-
Toward Intelligent Recommendations Using the Neural Knowledge DNA
PublikacjaIn this paper we propose a novel recommendation approach using past news click data and the Neural Knowledge DNA (NK-DNA). The Neural Knowledge DNA is a novel knowledge representation method designed to support discovering, storing, reusing, improving, and sharing knowledge among machines and computing systems. We examine our approach for news recommendation tasks on the MIND benchmark dataset. By taking advantages of NK-DNA, deep...
-
WYKORZYSTANIE SIECI NEURONOWYCH DO SYNTEZY MOWY WYRAŻAJĄCEJ EMOCJE
PublikacjaW niniejszym artykule przedstawiono analizę rozwiązań do rozpoznawania emocji opartych na mowie i możliwości ich wykorzystania w syntezie mowy z emocjami, wykorzystując do tego celu sieci neuronowe. Przedstawiono aktualne rozwiązania dotyczące rozpoznawania emocji w mowie i metod syntezy mowy za pomocą sieci neuronowych. Obecnie obserwuje się znaczny wzrost zainteresowania i wykorzystania uczenia głębokiego w aplikacjach związanych...
-
Awaria ściany szczelinowej stanowiącej obudowę wykopu głębokiego
PublikacjaW artykule przedstawiono awarię ściany szczelinowej stanowiącej obudowę wykopu głębokiego. Omówiono warunki geologiczne, hydrogeologiczne i geotechniczne podłoża. Przeprowadzono analizę rozwiązań konstrukcyjnych w kolejnych oszczędnościowych modyfikacjach projektu oraz zrealizowanej wersji. Wykonano obliczenia porównawcze. Wskazano przyczyny wystąpienia awarii. In the paper a failure of diaphragm wall protecting deep excavation...
-
Concurrent Video Denoising and Deblurring for Dynamic Scenes
PublikacjaDynamic scene video deblurring is a challenging task due to the spatially variant blur inflicted by independently moving objects and camera shakes. Recent deep learning works bypass the ill-posedness of explicitly deriving the blur kernel by learning pixel-to-pixel mappings, which is commonly enhanced by larger region awareness. This is a difficult yet simplified scenario because noise is neglected when it is omnipresent in a wide...
-
Zastosowanie mieszania wgłębnego i iniekcji strumieniowej do wzmocnienia podłoża pod budynkiem basenu.
PublikacjaPrzedstawiono jeden z pierwszych przykładów zastosowania w Europie nowoczesnej technologii wgłębnego mieszania gruntu na mokro (ang. wet Deep Soil Mixing) w celu posadowienia budynku basenu na uwarstwionym podłożu zawierającym grunt organiczny. Technologia DSM, wprowadzona w Polsce przez firmę Keller, polega na mechanicznym mieszaniu in situ gruntu z zaczynem cementowym i pozwala na formowanie w gruncie kolumn z cementogruntu w...
-
Routing Method for Interplanetary Satellite Communication in IoT Networks Based on IPv6
PublikacjaThe matter of interplanetary network (IPN) connection is a complex and sophisticated topic. Space missions are aimed inter alia at studying the outer planets of our solar system. Data transmission itself, as well as receiving data from satellites located on the borders of the solar system, was only possible thanks to the use of powerful deep space network (DSN) receivers, located in various places on the surface of the Earth. In...
-
Towards Knowledge Sharing Oriented Adaptive Control
PublikacjaIn this paper, we propose a knowledge sharing oriented approach to enable a robot to reuse other robots' knowledge by adapting itself to the inverse dynamics model of the knowledge-sharing robot. The purpose of this work is to remove the heavy fine-tuning procedure required before using a new robot for a task via reusing other robots' knowledge. We use the Neural Knowledge DNA (NK-DNA) to help robots gain empirical knowledge and...
-
Hybrid Processing by Turning and Burnishing of Machine Components
PublikacjaThe paper presents a method of hybrid manufacturing process of long 5 shafts and deep holes by simultaneous turning and burnishing method. The tech- 6 nological results of the research focus on the influence of the basic technological 7 parameters of this process on the surface roughness of piston rods of hydraulic 8 cylinders. Research results are presented in the graphs as well as mathematical 9 formula. Set of samples were made...
-
Analysis of the Application of Horizontal Directional Drilling
PublikacjaConstruction works are often considered to be very intrusive for the environment. Project designers assume deep excavations, or a complete replacement of the ground within the investment, which sometimes changes the initial conditions drastically. The problem started to appear in places, where the terrain is complicated and the excavation is burdensome. Some of state authorities...
-
Detailed studies of superconducting properties of Y2Pd1.25Ge2.75
PublikacjaWe report a successful synthesis of a high-purity intermetallic germanide Y2Pd1.25Ge2.75, crystallizing in the disordered variant of the AlB2-type structure. A single-phase sample was obtained via arc-melting by deliberately tuning the composition out of the ideal 2:1:3 ratio. Specific heat, electrical resistivity and magnetization measurements show that the compound is a weakly-coupled (λ e-p = 0.58) type-II superconductor with...
-
Towards a modification of a regulatory framework aiming at bunker oil spill prevention from ships - A design aspect of bunker tanks vents location guided by CFD simulations
PublikacjaAlthough accidental bunker oil spills at seaway are relatively rare events, they can be pose real threat to the natural marine environment. One of the reasons for the bunker spill to occur is a design failure; one of the ways it can demonstrate is an improper location or height of vent heads, leading to a bunker oil discharge during heavy rolling, due to sloshing phenomenon. Design of a ship and her systems is guided by various...
-
Comparison of the Ability of Neural Network Model and Humans to Detect a Cloned Voice
PublikacjaThe vulnerability of the speaker identity verification system to attacks using voice cloning was examined. The research project assumed creating a model for verifying the speaker’s identity based on voice biometrics and then testing its resistance to potential attacks using voice cloning. The Deep Speaker Neural Speaker Embedding System was trained, and the Real-Time Voice Cloning system was employed based on the SV2TTS, Tacotron,...
-
Grade of service determination methodology in IP networks with SIP protocol
PublikacjaAlthough Grade of Service is very important in VoIP providers evaluation, We wasn't able to find any paper regarding the topic of measuring GoS variables for IP networks utilizing SIP, which are defined like for PSTN/ISDN/GSM networks (post-selection delay, answering delay, release delay, or probability of end-to-end blocking). Due to the lack of research in this field, it was necessary to start from defining measures and cover...
-
Non-invasive blood glucose monitoring with Raman spectroscopy: prospects for device miniaturization
PublikacjaThe number of patients with diabetes has reached over 350 million, and still continues to increase. The need for regular blood glucose monitoring sparks the interest in the development of modern detection technologies. One of those methods, which allows for noninvasive measurements, is Raman spectroscopy. The ability of infrared light to penetrate deep into tissues allows for obtaining measurements through the skin without its...
-
Comparison of the exponential thermal transient parameterization methods with the SMTP method in the unipedicled DIEP flap computer modelling and simulation
PublikacjaThe aim of this paper is to compare the spatial contrast of the image descriptors obtained via three different thermal transient parameterization methods in Active Dynamic Thermography. The thermal constants and amplitude values of the one- and two- exponential parametrization are compared to the Simplified Magnitude-Temporal Parametrization method (SMTP). The comparison is performed using the data obtained by simulating the cold...
-
Method for Clustering of Brain Activity Data Derived from EEG Signals
PublikacjaA method for assessing separability of EEG signals associated with three classes of brain activity is proposed. The EEG signals are acquired from 23 subjects, gathered from a headset consisting of 14 electrodes. Data are processed by applying Discrete Wavelet Transform (DWT) for the signal analysis and an autoencoder neural network for the brain activity separation. Processing involves 74 wavelets from 3 DWT families: Coiflets,...
-
Experimental study on the selected aspects of bow thruster generated flow field at ship zero-speed conditions
PublikacjaThe paper presents the results of experimental study on the interaction between the bow thrusters understood as the flow field changes generated by bow tunnel thruster in deep water conditions operated as a single and twin units. The research was limited to zero-speed case for the ship dead in the water. The influence of the hull form and jet spread between the neighbouring thrusters for several combinations of the applied bow...
-
A general approach to study molecular fragmentation and energy redistribution after an ionizing event
PublikacjaWe propose to combine quantum chemical calculations, statistical mechanical methods, and photoionization and particle collision experiments to unravel the redistribution of internal energy of the furan cation and its dissociation pathways. This approach successfully reproduces the relative intensity of the different fragments as a function of the internal energy of the system in photoelectron–photoion coincidence experiments and...
-
From Linear Classifier to Convolutional Neural Network for Hand Pose Recognition
PublikacjaRecently gathered image datasets and the new capabilities of high-performance computing systems have allowed developing new artificial neural network models and training algorithms. Using the new machine learning models, computer vision tasks can be accomplished based on the raw values of image pixels instead of specific features. The principle of operation of deep neural networks resembles more and more what we believe to be happening...
-
Sensors and Sensor’s Fusion in Autonomous Vehicles
PublikacjaAutonomous vehicle navigation has been at the center of several major developments, both in civilian and defense applications. New technologies such as multisensory data fusion, big data processing, and deep learning are changing the quality of areas of applications, improving the sensors and systems used. New ideas such as 3D radar, 3D sonar, LiDAR, and others are based on autonomous vehicle revolutionary development. The Special...