Wyszukiwarka
Wyniki wyszukiwania dla: data augmentation
-
Toward Robust Pedestrian Detection With Data Augmentation
PublikacjaIn this article, the problem of creating a safe pedestrian detection model that can operate in the real world is tackled. While recent advances have led to significantly improved detection accuracy on various benchmarks, existing deep learning models are vulnerable to invisible to the human eye changes in the input image which raises concerns about its safety. A popular and simple technique for improving robustness is using data...
-
Musical Instrument Tagging Using Data Augmentation and Effective Noisy Data Processing
PublikacjaDeveloping signal processing methods to extract information automatically has potential in several applications, for example searching for multimedia based on its audio content, making context-aware mobile applications (e.g., tuning apps), or pre-processing for an automatic mixing system. However, the last-mentioned application needs a significant amount of research to reliably recognize real musical instruments in recordings....
-
Data augmentation for improving deep learning in image classification problem
PublikacjaThese days deep learning is the fastest-growing field in the field of Machine Learning (ML) and Deep Neural Networks (DNN). Among many of DNN structures, the Convolutional Neural Networks (CNN) are currently the main tool used for the image analysis and classification purposes. Although great achievements and perspectives, deep neural networks and accompanying learning algorithms have some relevant challenges to tackle. In this...
-
Learning sperm cells part segmentation with class-specific data augmentation
PublikacjaInfertility affects around 15% of couples worldwide. Male fertility problems include poor sperm quality and low sperm count. The advanced fertility treatment methods like ICSI are nowadays supported by vision systems to assist embryologists in selecting good quality sperm. Computer-Assisted Semen Analysis (CASA) provides quantitative and qualitative sperm analysis concerning concentration, motility, morphology, vitality, and fragmentation....
-
Detection of Lexical Stress Errors in Non-Native (L2) English with Data Augmentation and Attention
PublikacjaThis paper describes two novel complementary techniques that improve the detection of lexical stress errors in non-native (L2) English speech: attention-based feature extraction and data augmentation based on Neural Text-To-Speech (TTS). In a classical approach, audio features are usually extracted from fixed regions of speech such as the syllable nucleus. We propose an attention-based deep learning model that automatically de...
-
Application Of Generative Adversarial Network for Data Augmentation and Multiplication to Automated Cell Segmentation of the Corneal Endothelium
PublikacjaConsidering the automatic segmentation of the endothelial layer, the available data of the corneal endothelium is still limited to a few datasets, typically containing an average of only about 30 images. To fill this gap, this paper introduces the use of Generative Adversarial Networks (GANs) to augment and multiply data. By using the ``Alizarine'' dataset, we train a model to generate a new synthetic dataset with over 513k images....
-
BETWEEN IDEA AND INTERPRETATION - DESIGN PROCESS AUGMENTATION
PublikacjaThe following paper investigates the idea of reducing the human digital intervention to a minimum during the advanced design process. Augmenting the outcome attributes beyond the designer's capabilities by computational design methods, data collection, data computing and digital fabrication, altogether imitating the human design process. The primary technical goal of the research was verification of restrictions and abilities used...
-
Universal Augmentation Schemes for Network Navigability
PublikacjaRozważano problem uzupełniania grafu (reprezentującego np. sieci społeczne) poprzez dodanie w każdym węźle jednego dodatkowego skierowanego połączenia (długodystansowego). Dokładniej, dla każdego węzła definiuje się listę prawdopodobieństw istnienia połączenia wychodzącego z danego węzła do wszystkich pozostałych węzłów; wartości tych prawdopodobieństw muszą sumować się do jedności. Routing zachłanny w takiej sieci polega na przekazywaniu...
-
Intelligence Augmentation and Amplification: Approaches, Tools, and Case Studies
PublikacjaMost experts agree that truly intelligent artificial system is yet to be developed. The main issue that still remains a challenge is imposing trust and explainability into such systems. However, is full replication of human intelligence really desirable key aim in intelligence related technology and research? This is where the concept of augmented intelligence comes into play. It is an alternative conceptualization of artificial...
-
Heat transfer augmentation using liquid crystal
PublikacjaPraca przedstawia zastosowanie techniki wizualizacji ciekłokrystalicznej dobadań wymiany ciepła na płaskiej powierzchni z zaburzaczami. Folia ciekło-krystaliczna została umieszczona na powierzchni, aby określić przebieg izoterm. Obraz rejestrowano kamerą JVC, a następnie analizowano komputerowo. Procedura eksperymentalna pozwala na otrzymanie pełnej informacji dotyczącej wymiany ciepła (na powierzchni płaskiej z elementami ustawionymi...
-
The Optimal Location of Ground-Based GNSS Augmentation Transceivers
PublikacjaModern Global Navigation Satellite Systems (GNSS) allow for positioning with accuracies ranging from tens of meters to single millimeters depending on user requirements and available equipment. A major disadvantage of these systems is their unavailability or limited availability when the sky is obstructed. One solution is to use additional range measurements from ground-based nodes located in the vicinity of the receiver. The highest...
-
Experience-based Intelligence Augmentation with Decisional DNA: Upcoming direction
PublikacjaIntelligence amplification systems and technologies have gained significant interest from academia and industry during the past few decades. One of the main reasons behind this trend is the fact that most experts agree that truly intelligent artificial system is yet to be developed. The question increasing often asked is this: Is full replication of human intelligence desirable key aim in intelligence related technology and research?...
-
Heat transfer augmentation visualized by liquid crystal thermography
PublikacjaW pracy przedstawiono możliwości wykorzystania termografii ciekłokrystalicznej do badań wymiany ciepła oraz wizualizacji pola przepływu w elementach urządzeń przemysłowych.
-
Toughness augmentation by fibrillation and yielding in nanostructured blends with recycled polyurethane as a modifier
PublikacjaIn the present paper, we have carefully investigated the morphology and fracture mechanism of the recycled polyurethane (RPU) /epoxy blend system. The second phase (RPU) added to the epoxy resin has a positive effect on the overall mechanical properties. Interestingly, the recycled polymer has a remarkable effect on the fracture toughness of epoxy resin. The mechanism behind the fracture toughness improvement up on the addition...
-
Possibilities of heat transfer augmentation in heat exchangers with minichannels for marine applications
PublikacjaIn the paper, new trends in development of microchannel heat exchangers are presented. The exchangers developed in this way can be applied in marine applications. Main attention was focused on heat exchangers design with reduced size of passages namely based on microchannels. In our opinion future development of high power heat exchangers will be based on networks micro heat exchangers.
-
The (de)biasing Effect of GAN-Based Augmentation Methods on Skin Lesion Images
Publikacja -
Power augmentation of PGE Gorzow's gas turbine by steam injection - thermodynamic overview
PublikacjaThe impact of steam injection upon a gas turbine and a combined power plant performance has been investigated. This article describes and summarizes possibilities of modification for current gas turbine in PGE Gorzow power plant into the Cheng cycle. Our modification deals with a thermal cycle, in which steam produced in a heat recovery steam generator is injected into the gas turbine's combustion chamber. It has been proved that...
-
Changes of augmentation index early after ischaemic stroke predict functional outcome
Publikacja -
Cerclage cable augmentation does not increase stability of the fixation of intertrochanteric fractures. A biomechanical study
PublikacjaBackground: Intertrochanteric fractures with a posteromedial intermediate fragment are unstable because of the loss of medial support. Additional fixation with a cerclage is used in subtrochanteric fractures, but not in intertrochanteric fractures. The aim of this biomechanical study is to evaluate whether cerclage fixation improves stability of intertrochanteric fractures. Hypothesis: Our hypothesis is that the cerclage fixation...
-
Heat Transfer Augmentation through Different Jet Impingement Techniques: A State-of-the-Art Review
Publikacja -
AUGMENTATION OF THE CRITICAL HEAT FLUX IN WATER-Al2O3, WATER-TiO2 AND WATER-Cu NANOFLUIDS
PublikacjaThe main aim of the proposed study is therefore recognition of the phenomena accompanying nucleate boiling crisis of selected nanofluids during boiling on horizontal tubes of various outside diameters. Of particular interest is impact of contact angle and tube diameter on the value of critical heat flux. The results obtained should give more light on the nature of nucleate boiling crisis and will serve as a basis for future theoretical...
-
Contribution of arterial excess pressure and arterial stiffness to central augmentation pressure in healthy subjects
Publikacja -
Mobilenet-V2 Enhanced Parkinson's Disease Prediction with Hybrid Data Integration
PublikacjaThis study investigates the role of deep learning models, particularly MobileNet-v2, in Parkinson's Disease (PD) detection through handwriting spiral analysis. Handwriting difficulties often signal early signs of PD, necessitating early detection tools due to potential impacts on patients' work capacities. The study utilizes a three-fold approach, including data augmentation, algorithm development for simulated PD image datasets,...
-
Augmentation index, pulse pressure amplification and superoxide anion production in patients with coronary artery disease
Publikacja -
Genetic and pharmacologic proteasome augmentation ameliorates Alzheimer’s-like pathology in mouse and fly APP overexpression models
PublikacjaThe proteasome has key roles in neuronal proteostasis, including the removal of misfolded and oxidized proteins, presynaptic protein turnover, and synaptic efficacy and plasticity. Proteasome dysfunction is a prominent feature of Alzheimer’s disease (AD). We show that prevention of proteasome dysfunction by genetic manipulation delays mortality, cell death, and cognitive deficits in fly and cell culture AD models. We developed...
-
Genetic and pharmacologic proteasome augmentation ameliorates Alzheimer’s-like pathology in mouse and fly APP overexpression models
PublikacjaThe proteasome has key roles in neuronal proteostasis, including the removal of misfolded and oxidized proteins, presynaptic protein turnover, and synaptic efficacy and plasticity. Proteasome dysfunction is a prominent feature of Alzheimer’s disease (AD). We show that prevention of proteasome dysfunction by genetic manipulation delays mortality, cell death, and cognitive deficits in fly and cell culture AD models. We developed...
-
Categorization of emotions in dog behavior based on the deep neural network
PublikacjaThe aim of this article is to present a neural system based on stock architecture for recognizing emotional behavior in dogs. Our considerations are inspired by the original work of Franzoni et al. on recognizing dog emotions. An appropriate set of photographic data has been compiled taking into account five classes of emotional behavior in dogs of one breed, including joy, anger, licking, yawning, and sleeping. Focusing on a particular...
-
Closer Look at the Uncertainty Estimation in Semantic Segmentation under Distributional Shift
PublikacjaWhile recent computer vision algorithms achieve impressive performance on many benchmarks, they lack robustness - presented with an image from a different distribution, (e.g. weather or lighting conditions not considered during training), they may produce an erroneous prediction. Therefore, it is desired that such a model will be able to reliably predict its confidence measure. In this work, uncertainty estimation for the task...
-
Spatial Visualization Based on Geodata Fusion Using an Autonomous Unmanned Vessel
PublikacjaThe visualization of riverbeds and surface facilities on the banks is crucial for systems that analyze conditions, safety, and changes in this environment. Hence, in this paper, we propose collecting, and processing data from a variety of sensors—sonar, LiDAR, multibeam echosounder (MBES), and camera—to create a visualization for further analysis. For this purpose, we took measurements from sensors installed on an autonomous, unmanned...
-
Robustness in Compressed Neural Networks for Object Detection
PublikacjaModel compression techniques allow to significantly reduce the computational cost associated with data processing by deep neural networks with only a minor decrease in average accuracy. Simultaneously, reducing the model size may have a large effect on noisy cases or objects belonging to less frequent classes. It is a crucial problem from the perspective of the models' safety, especially for object detection in the autonomous driving...
-
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)...
-
Repair Augmentation of Unstable, Complete Vertical Meniscal Tears With Bone Marrow Venting Procedure: A Prospective, Randomized, Double-Blind, Parallel-Group, Placebo-Controlled Study
Publikacja -
Detecting Lombard Speech Using Deep Learning Approach
PublikacjaRobust Lombard speech-in-noise detecting is challenging. This study proposes a strategy to detect Lombard speech using a machine learning approach for applications such as public address systems that work in near real time. The paper starts with the background concerning the Lombard effect. Then, assumptions of the work performed for Lombard speech detection are outlined. The framework proposed combines convolutional neural networks...
-
Agnieszka Mikołajczyk-Bareła dr inż.
Osoby -
Author Reply to “Regarding ‘Repair Augmentation of Unstable, Complete Vertical Meniscal Tears With Bone Marrow Venting Procedure: A Prospective, Randomized, Double-Blind, Parallel-Group, Placebo-Controlled Study’”
Publikacja -
Bias mitigation benchmark that includes two datasets
Dane BadawczeISIC-2020 is the largest skin lesion dataset divided into two classes -- benign and malignant. It contains 33126 dermoscopic images from over 2000 patients. The diagnoses were confirmed either by histopathology, expert agreement or longitudinal follow-up. The dataset was gathered by The International Skin Imaging Collaboration (ISIC) from several medical...
-
Investigating Feature Spaces for Isolated Word Recognition
PublikacjaThe study addresses the issues related to the appropriateness of a two-dimensional representation of speech signal for speech recognition tasks based on deep learning techniques. The approach combines Convolutional Neural Networks (CNNs) and time-frequency signal representation converted to the investigated feature spaces. In particular, waveforms and fractal dimension features of the signal were chosen for the time domain, and...
-
Towards Precise Visual Navigation and Direct Georeferencing for MAV Using ORB-SLAM2
PublikacjaA low accuracy of positioning using Global Navigation Satellite System (GNSS) are not meet geodetic requirements for direct images georeferencing for Unmanned Aerial Vehicle (UAV) photogrammetry. A majority of UAVs are equipped with a monocular or stereo non-metric cameras for either visual data gathering or live video feed for operator. A cheap positioning techniques used on board commercial UAVs are not that precise as geodetic...
-
High precision and accuracy using low cost GNSS receivers and supporting technologies
PublikacjaThis chapter focuses on methods and techniques of positioning, based on the highly accurate and precise Global Navigation Satellite System (GNSS), which are available at a relatively low price. In this context, a comparison of different positioning methods provided by the free and open source software (FOSS) package called RTKLIB is given. Other aspects related to price reduction are also considered, including availability and...
-
Clothes Detection and Classification Using Convolutional Neural Networks
PublikacjaIn this paper we describe development of a computer vision system for accurate detection and classification of clothes for e-commerce images. We present a set of experiments on well established architectures of convolutional neural networks, including Residual networks, SqueezeNet and Single Shot MultiBox Detector (SSD). The clothes detection network was trained and tested on DeepFashion dataset, which contains box annotations...
-
A CNN based coronavirus disease prediction system for chest X-rays
PublikacjaCoronavirus disease (COVID-19) proliferated globally in early 2020, causing existential dread in the whole world. Radiography is crucial in the clinical staging and diagnosis of COVID-19 and offers high potential to improve healthcare plans for tackling the pandemic. However high variations in infection characteristics and low contrast between normal and infected regions pose great challenges in preparing radiological reports....
-
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...
-
Convolutional Neural Networks for C. Elegans Muscle Age Classification Using Only Self-Learned Features
PublikacjaNematodes Caenorhabditis elegans (C. elegans) have been used as model organisms in a wide variety of biological studies, especially those intended to obtain a better understanding of aging and age-associated diseases. This paper focuses on automating the analysis of C. elegans imagery to classify the muscle age of nematodes based on the known and well established IICBU dataset. Unlike many modern classification methods, the proposed...
-
Computer-Aided Detection of Hypertensive Retinopathy Using Depth-Wise Separable CNN
PublikacjaHypertensive retinopathy (HR) is a retinal disorder, linked to high blood pressure. The incidence of HR-eye illness is directly related to the severity and duration of hypertension. It is critical to identify and analyze HR at an early stage to avoid blindness. There are presently only a few computer-aided systems (CADx) designed to recognize HR. Instead, those systems concentrated on collecting features from many retinopathy-related...
-
Full CI ground state potential energy curves and one-electron relativistic corrections for hydrogen molecule in various basis sets
Dane BadawczeThis dataset consists of Full CI ground state Born-Oppenheimer potential energy curves and one-electron relativistic corrections for hydrogen dimer. Nonrelativistic energies, as well as one electron relativistic corrections (treated perturbatively with help of the Cowan-Griffin Hamiltonian) are presented for internuclear distances between 0.8 and 10...
-
A Parallax Shift Effect Correction Based on Cloud Height for Geostationary Satellites and Radar Observations
PublikacjaThe effect of cloud parallax shift occurs in satellite imaging, particularly for high angles of satellite observations. This study demonstrates new methods of parallax effect correction for clouds observed by geostationary satellites. The analytical method that could be found in literature, namely the Vicente et al./Koenig method, is presented at the beginning. It approximates a cloud position using an ellipsoid with semi-axes...
-
Has EGNOS its own place in maritime navigation?
PublikacjaGNSS (in fact it usually means GPS) is at the moment the fundamental means of positioning for any variants of navigation. However some attributes of it can be not sufficient for every users or in some tasks, so there are any cases when this system needs support. From the mariner point of view pure GPS properties perfectly fulfils expectations of sailors at the open sea, however there are lot of cases, when accuracy of the system...
-
Evaluating calibration and robustness of pedestrian detectors
PublikacjaIn this work robustness and calibration of modern pedestrian detectors are evaluated. Pedestrian detection is a crucial perception com- ponent in autonomous driving and here we study its performance under different image corruptions. Furthermore, we provide analysis of classifi- cation calibration of pedestrian detectors and we show a positive effect of using style-transfer augmentation technique. Our analysis is aimed as a step...
-
AUTOMATED NEGOTIATIONS OVER COLLABORATION PROTOCOL AGREEMENTS
PublikacjaThe dissertation focuses on the augmentation of proactive document - agents with built-in intelligence to recognize execution context provided by devices visited during a business process, and to reach collaboration agreement despite conflicting requirements. The proposed solution, based on intelligent bargaining using neural networks to improve simple multi-issue negotiation between the document and thedevice, requires practically...
-
Representing and Managing Experiential Knowledge with Decisional DNA and its Drimos® Extension
PublikacjaThe Semantic Web concept is proposing a future concept of the WorldWideWeb (WWW) where both humans and man-made systems are able to interconnect and exchange knowledge. One of the challenges of Semantic Web is smart and trusted accommodation of knowledge in artificial systems so it can be unified, enhanced, reused, shared, communicated and distributed with added aptitude. Our research represents an important component of addressing...