A5088524211- Fetal and Pediatric Neurological Disorders
- Aerospace, Electronics, Mathematical Modeling
- Advanced Image Processing Techniques
- Advanced Neuroimaging Techniques and Applications
- Advanced MRI Techniques and Applications
- Medical Imaging Techniques and Applications
- Generative Adversarial Networks and Image Synthesis
- Neonatal and fetal brain pathology
- Diamond and Carbon-based Materials Research
- Advanced Measurement and Detection Methods
- Aesthetic Perception and Analysis
- Metal and Thin Film Mechanics
- Image Retrieval and Classification Techniques
- Advanced Theoretical and Applied Studies in Material Sciences and Geometry
- Image Enhancement Techniques
- Heat Transfer and Mathematical Modeling
- Rocket and propulsion systems research
- Technical Engine Diagnostics and Monitoring
- Advanced Research in Systems and Signal Processing
- Advanced Neural Network Applications
- MRI in cancer diagnosis
- Engine and Fuel Emissions
- Aerospace Engineering and Control Systems
- Online Learning and Analytics
- Image and Video Quality Assessment
King's College London
2023-2025
Sukhoi State Technical University of Gomel
2025
Moscow Aviation Institute
2020-2022
Baikov Institute of Metallurgy and Materials Science
2021
Skolkovo Institute of Science and Technology
2019
The paper presents a designed model of maintenance repair and overhaul (MRO) an aircraft with predictive unit. use artificial neural network is provided as tool. Predictive part the MRO strategy "Before fault detected". project will allow to build concept conducting in-depth analysis before choosing for controlling stage life cycle - operation.
Image Quality Assessment (IQA) metrics are widely used to quantitatively estimate the extent of image degradation following some forming, restoring, transforming, or enhancing algorithms. We present PyTorch (PIQ), a usability-centric library that contains most pop- ular modern IQA algorithms, guaranteed be correctly implemented according their original propositions and thoroughly verified. In this paper, we detail principles behind founda- tion library, describe evaluation strategy makes it...
The article develops a typical structural diagram of the hydraulic system short-haul passenger aircraft, obtained on basis an analysis diagrams aircraft this class. This made it possible to carry out generalized calculation required energy characteristics consumers and as whole for modes: takeoff, flight landing, well develop kinematic calculate cylinder-lift front landing gear. Also in article, verification cylinder stability supportive position rod compression was carried out.
Abstract Recent advances in fetal fMRI present a new opportunity for neuroscience to study functional human brain connectivity at the time of its emergence. Progress field however has been hampered by lack openly available datasets that can be exploited researchers across disciplines develop methods would address unique challenges associated with imaging and analysing utero, such as unconstrained head motion, dynamically evolving geometric distortions, or inherently low signal-to-noise...
This paper explores the application of machine learning methods for recognizing automobile light signals to enhance smart traffic systems. For vehicle detection in video footage, Keras library was employed along with RetinaNet neural network architecture [1]. The YOLOv8 used identifying status headlights and taillights. Data collection, annotation, model training were conducted using Roboflow platform. research resulted trained weights capable state front rear lights on various types under...
Image Quality Assessment (IQA) metrics are widely used to quantitatively estimate the extent of image degradation following some forming, restoring, transforming, or enhancing algorithms. We present PyTorch (PIQ), a usability-centric library that contains most popular modern IQA algorithms, guaranteed be correctly implemented according their original propositions and thoroughly verified. In this paper, we detail principles behind foundation library, describe evaluation strategy makes it...
Abstract The effect of high-pressure synthesis temperature on the structure and indentation characteristics superelastic hard carbon formed from amorphous fullerites tribological properties Co-based composite materials (CM) reinforced by particles such has been studied Raman spectroscopy, high-resolution transmission electron microscopy (HRTEM), measurements, tests. It is shown that ball milling (for 48 h) C 60 fullerite crystals results in amorphization product transformation upon their...
The paper considers the issues of increasing energy efficiency extinguishing forest fires by improving equipment installed on transport aircraft, in particular, improvement individual modules existing VAP-2 aviation pouring device, which is widely used Il–76 military aircraft. Increasing fire achieved automating process water discharge, for a simulation model has been developed that allows you to calculate when, at what expense and amount time carry out discharge as efficiently possible,...
This paper introduces a new data-driven, non-parametric method for image quality and aesthetics assessment, surpassing existing approaches requiring no prompt engineering or fine-tuning. We eliminate the need expressive textual embeddings by proposing efficient anchors in data. Through extensive evaluations of 7 state-of-the-art self-supervised models, our demonstrates superior performance robustness across various datasets benchmarks. Notably, it achieves high agreement with human...
This paper examined the problem of noise in aircraft cabin and ways to solve it. Through observations, following causes were identified: power plant (acoustic mechanical vibrations, turbulent boundary layer others), location engine relative airframe, propeller (if it is a plant), itself, air conditioning system, so on same operating device cabin. There comprehensive way cabin: for with turbojet engines, measures aimed at changing design use sound- absorbing materials can be effective. For...
Abstract Recent advances in fetal fMRI present a new opportunity for neuroscience to study functional human brain connectivity at the time of its emergence. Progress field however has been hampered by lack openly available datasets that can be exploited researchers across disciplines develop methods would address unique challenges associated with imaging and analysing utero, such as unconstrained head motion, dynamically evolving geometric distortions, or inherently low signal-to-noise...
Dynamic free-breathing fetal heart MRI requires high spatial and temporal resolution, which could be reconstructed by kt-SENSE from undersampled data guided priors of the same anatomy. Doubled acquisition time uncontrolled motion between 2 acquisitions affects quality for reconstruction. We explored an alternative deep learning approach using a 3D U-Net based model with time-averaged skip connection consistency. Assessment set baseline prior preconstruction underlines important pitfalls that...
Dynamic fetal heart magnetic resonance imaging (MRI) presents unique challenges due to the fast rate of fetus compared adult subjects and uncontrolled motion. This requires high temporal spatial resolutions over a large field view, in order encompass surrounding maternal anatomy. In this work, we introduce Cardiac Reconstruction Attention Network (DCRA-Net) - novel deep learning model that employs attention mechanisms domains frequency representation data reconstruct dynamics from highly...
The composite materials (CM) based on Co, Ti-Cu, and Pt reinforced with diamond-like carbon particles diamonds different component ratios have been obtained by high-pressure high-temperature synthesis (5–9 GPa, 800–1800 °C). methods of optical microscopy, dynamic indentation, tribological tests were used to establish the correlation between structure properties reinforcing phase CM define limits applicability as tribotechnical materials. Low friction coefficients (0.08–0.06) demonstrated a...
Dynamic free-breathing fetal cardiac MRI is one of the most challenging modalities, which requires high temporal and spatial resolution to depict rapid changes in a small heart. The ability deep learning methods recover undersampled data could help optimise kt-SENSE acquisition strategy improve non-gated reconstruction quality. In this work, we explore supervised networks for style acquired using an extensive vivo dataset. Having access fully-sampled low-resolution multi-coil MRI, study...
Synthetic CT image generation from MRI scan is necessary to create radiotherapy plans without the need of co-registered and scans. The chosen baseline adversarial model with cycle consistency permits unpaired image-to-image translation. Perceptual loss function term coordinate convolutional layer were added improve quality translated images. proposed architecture was tested on paired MRI-CT dataset, where synthetic CTs compared corresponding original MAE between images real scans 61 HU...
While studying at a higher educational institution, student is immersed in unique world with its own rules and laws. Training various forms of specialization lasts from 4 to 5.5 years. During this time, the manages replenish his knowledge base competencies qualities that will later help achieve successful career growth. It obvious getting education plays significant role formation personality, schools have everything necessary test acquired skills. The flexibility curricula allows students...