VNU-UET Repository: No conditions. Results ordered -Date Deposited. 2023-03-26T14:40:41ZEPrintshttp://eprints.uet.vnu.edu.vn/images/sitelogo.pnghttps://eprints.uet.vnu.edu.vn/eprints/2022-08-22T04:06:31Z2022-08-22T04:06:31Zhttp://eprints.uet.vnu.edu.vn/eprints/id/eprint/4777This item is in the repository with the URL: http://eprints.uet.vnu.edu.vn/eprints/id/eprint/47772022-08-22T04:06:31ZĐánh Giá Hiệu Năng Một Số Kỹ Thuật Học Sâu Cho Phân Vùng Mạch Máu Gan Trong Ảnh Chụp Cắt Lớp Vi TínhPhân vùng mạch máu gan trong ảnh chụp cắt lớp vi tính (CLVT) là bước quan trọng đối với việc chẩn đoán và lập kế hoạch điều trị các bệnh lý liên quan tới gan. Mạch máu gan có cấu trúc phức tạp, độ tương phản thấp, ảnh hưởng bởi nhiễu, do đó đã đặt ra nhiều thách thức cho việc phân vùng chính xác mạch máu gan. Trong lĩnh vực xử lý ảnh y tế, kỹ thuật học sâu đã cho thấy sự phát triển nhanh chóng. Gần đây, mạng nơ-ron Transformer được áp dụng và cho kết quả khả quan trong lĩnh vực xử lý ảnh y tế. Trong bài báo này, chúng tôi đánh giá hiệu năng phân vùng mạch máu gan giữa kỹ thuật học sâu dựa trên mạng nơ-ron tích chập (3D-ResUNet, 2D/3D nn-UNet) và kỹ thuật học sâu sử dụng mạng nơ-ron Transformer (TransUNet, Swin-UNet, MedT). Dữ liệu ảnh chụp CLVT sử dụng để huấn luyện và đánh giá được thu thập từ nhiều cơ sở y tế trên thế giới bao gồm ảnh chụp CLVT sử dụng bức xạ liều thông thường và liều thấp. Kết quả cho thấy 3D nn-UNet có độ chính xác (ACC) trung bình cao nhất, 98\%; ba kỹ thuật học sâu 2D/3D nn-UNet và TransUNet đều đạt trung bình giá trị chỉ số đánh giá DSC lớn hơn 75\%. Trong kỹ thuật điều trị đốt sóng cao tần (RFA) mạch máu gan lớn là vùng được quan tâm, cả ba kỹ thuật học sâu nêu ở trên đều cho chỉ số đánh giá trung bình DSC ở vùng mạch máu lớn lớn hơn 80\%. Kết quả của nghiên cứu cho thấy rằng 3D nn-UNet có thể tự động phân vùng mạch máu gan với độ chính xác cao, cho thấy tiềm năng ứng dụng vào quá trình lập kế hoạch can thiệp điều trị ung thư gan bằng kỹ thuật RFA.Quoc Anh Lequocanh.uet@gmail.comXuan Loc Phamxuanloc97ars@vnu.edu.vnManh Ha Luuhalm@vnu.edu.vn2022-03-21T00:46:35Z2022-03-21T00:46:35Zhttp://eprints.uet.vnu.edu.vn/eprints/id/eprint/4711This item is in the repository with the URL: http://eprints.uet.vnu.edu.vn/eprints/id/eprint/47112022-03-21T00:46:35ZMisspecified Cramer–Rao Bounds for Blind Channel Estimation Under Channel Order MisspecificationIn estimation, the misspecified Cramer–Rao bound (MCRB), which is an extension of the well-known Cramer–Rao bound (CRB) when the underlying system model is misspecified, has recently attracted much attention. In this paper, we introduce a new interpretation of the MCRB, called the generalized MCRB (GMCRB), via the Moore–Penrose inverse operator. This bound is useful for singular problems and particularly blind channel estimation problems in which the Hessian matrix is noninvertible. Two closed-form expressions of the GMCRB are derived for unbiased blind estimators when the channel order is misspecified. The first bound deals with deterministic models where both the channel and unknown symbols are deterministic. The second one is devoted to stochastic models where we assume that transmitted symbols are unknown random variables i.i.d. drawn from a Gaussian distribution. Two case studies of channel order misspecification are investigated to demonstrate the effectiveness of the proposed GMCRBs over the classical CRBs. When the channel order is known or accurately estimated, both generalized bounds reduce to the classical bounds. Besides, the stochastic GMCRB is lower than the deterministic one, especially at high SNR.Trung Thanh Leletrungthanhtbt@gmail.comAbed Meraim Karimkarim.abed-meraim@univ-orleans.frLinh Trung Nguyenlinhtrung@vnu.edu.vn2022-03-21T00:43:46Z2022-03-21T00:43:46Zhttp://eprints.uet.vnu.edu.vn/eprints/id/eprint/4703This item is in the repository with the URL: http://eprints.uet.vnu.edu.vn/eprints/id/eprint/47032022-03-21T00:43:46ZAn Effective Framework of Private Ethereum Blockchain Networks for Smart GridA smart grid is an important application in Industry 4.0 with a lot of new technologies and equipment working together. Hence, sensitive data stored in the smart grid is vulnerable to malicious modification and theft. This paper proposes a framework to build a smart grid based on a highly effective private Ethereum network. Our framework provides a real smart grid that includes modern hardware and a smart contract to secure data in the blockchain network. To obtain high throughput but a low uncle rate, the difficulty calculation method used in the mining process of the Ethereum consensus mechanism is modified to adapt to the practical smart grid setup. The performance in terms of throughput and latency are evaluated by simulation and verified by the real smart grid setup. The enhanced private Ethereum-based smart grid has significantly better performance than the public one. Moreover, this framework can be applied to any system used to store data in the Ethereum network.Do Hai SonTran Thi Thuy QuynhTran Viet KhoaDinh Thai HoangNguyen Linh TrungNguyen Viet HaDusit NiyatoDiep N. NguyenEryk Dutkiewicz2022-03-21T00:43:26Z2022-03-21T00:43:26Zhttp://eprints.uet.vnu.edu.vn/eprints/id/eprint/4704This item is in the repository with the URL: http://eprints.uet.vnu.edu.vn/eprints/id/eprint/47042022-03-21T00:43:26ZĐồng bộ nhiều SDR trong thực thi thuật toán ước lượng hướng sóng đến MUSICCác thiết bị SDR (vô tuyến định nghĩa bằng phần mềm) được sử dụng nhiều trong các lĩnh vực như vô tuyến điện nghiệp dư, thử nghiệm các hệ thống truyền thông hay giáo dục. Bài báo trình bày giải pháp đồng bộ các thiết bị SDR riêng lẻ có thể sử dụng cho các hệ thống xử lý tín hiệu mảng. Hệ thống được thử nghiệm dựa trên việc thực thi thời gian thực thuật toán ước lượng hướng sóng đến MUSIC với cả với tín hiệu băng hẹp và tín hiệu băng rộng. Đóng góp của bài báo bao gồm phương pháp và các khối được lập trình trên phần mềm GNU Radio
nhằm đồng bộ hóa mảng thu và ước lượng hướng sóng đến. Các kết quả trong bài báo cho thấy, việc đồng bộ và ước lượng hướng sóng đến trên hệ nhiều thiết bị SDR là hoàn toàn khả thi.Son Do Haidohaison1998@vnu.edu.vnQuynh Tran Thi Thuyquynhttt@vnu.edu.vn2022-03-21T00:42:04Z2022-03-21T00:42:04Zhttp://eprints.uet.vnu.edu.vn/eprints/id/eprint/4697This item is in the repository with the URL: http://eprints.uet.vnu.edu.vn/eprints/id/eprint/46972022-03-21T00:42:04ZSupSLAM: A Robust Visual Inertial SLAM System Using SuperPoint for Unmanned Aerial VehiclesSimultaneous localization and mapping (SLAM)
is essential for unmanned aerial vehicle (UAV) applications
since it allows the UAV to estimate not only its position and
orientation but also the map of its working environment. We
propose in this study a new SLAM system for UAVs named
SupSLAM that works with a stereo camera and an inertial
measurement unit (IMU). The system includes a front-end that
provides an initial estimate of the UAV position and working
environment and a back-end that compensates for the drift
caused by the initial estimation. To improve the accuracy and
robustness of the system, we use a new feature extraction
method named SuperPoint, which includes a pretrained deep
neural network to detect key points for estimation. This method is not only accurate in feature extraction but also efficient in computation so that it is relevant to implement on UAVs. We have conducted a number of experiments and comparisons to evaluate the performance of the proposed system. The results show that the system is feasible for UAV SLAM with the performance comparable to state-of-art methods in most datasets and better in some challenging conditions.Cong Hoang Quachhoangqc@vnu.edu.vnManh Duong Phungduongpm@vnu.edu.vnVu Ha Lehalv@vnu.edu.vnStuart PerryStuart.Perry@uts.edu.au2021-11-09T09:38:01Z2021-12-09T09:37:19Zhttp://eprints.uet.vnu.edu.vn/eprints/id/eprint/4635This item is in the repository with the URL: http://eprints.uet.vnu.edu.vn/eprints/id/eprint/46352021-11-09T09:38:01ZTowards Vision-based Concrete Crack Detection: Automatic Simulation of Real-world CracksRecently, automatic pavement crack detection using computer vision has attracted great attention from many research groups globally. One of the major concerns in this field, especially for deep learning-based approaches, is data collection, where human resources, time, and space are essential factors. This process remains a challenging task and significantly influences the results of crack detection algorithms, particularly when applied to large infrastructures such as bridges, dams, and high-rise buildings. This study aims to combine the synthesis of knowledge from fracture mechanics, simulation, machine learning, and research on concrete crack propagation to simulate crack propagation. Based on the simulation results, mathematical models of simulated crack will be generated using machine learning. Experimental results have been conducted on various reputable crack image datasets, showing a high correlation between the simulated cracks and the real-world ones. This novel, interdisciplinary approach is expected to have many useful points for a research area where data collection and labeling are still facing many difficulties.Tran Hiep Dinhtranhiep.dinh@vnu.edu.vnThi Thuy Anh Vuanhvutt@vnu.edu.vnTruongGiang Nguyenntgiang@imech.vast.vnCong Hieu Le18020508@vnu.edu.vnLinh Trung Nguyenlinhtrung@vnu.edu.vnNguyen Dinh Ducducnd@vnu.edu.vnChin-Teng Linchin-teng.lin@uts.edu.au2021-10-31T00:44:38Z2021-10-31T00:44:38Zhttp://eprints.uet.vnu.edu.vn/eprints/id/eprint/4622This item is in the repository with the URL: http://eprints.uet.vnu.edu.vn/eprints/id/eprint/46222021-10-31T00:44:38ZIterative Learning Sliding Mode Control for UAV Trajectory TrackingThis paper presents a novel iterative learning sliding mode controller (ILSMC) that can be applied to the trajectory tracking of quadrotor unmanned aerial vehicles (UAVs) subject to model uncertainties and external disturbances. Here, the proposed ILSMC is integrated in the outer loop of a controlled system. The control development, conducted in the discrete-time domain, does not require a priori information of the disturbance bound as with conventional SMC techniques. It only involves an equivalent control term for the desired dynamics in the closed loop and an iterative learning term to drive the system state toward the sliding surface to maintain robust performance. By learning from previous iterations, the ILSMC can yield very accurate tracking performance when a sliding mode is induced without control chattering. The design is then applied to the attitude control of a 3DR Solo UAV with a built-in PID controller. The simulation results and experimental validation with real-time data demonstrate the advantages of the proposed control scheme over existing techniques.Van Lanh NguyenManh Duong Phungduongpm@vnu.edu.vnHa Quangquang.ha@uts.edu.au2021-07-06T02:46:57Z2021-07-06T02:46:57Zhttp://eprints.uet.vnu.edu.vn/eprints/id/eprint/4561This item is in the repository with the URL: http://eprints.uet.vnu.edu.vn/eprints/id/eprint/45612021-07-06T02:46:57ZRobust Subspace Tracking Algorithms in Signal Processing: A Brief SurveyPrincipal component analysis (PCA) and subspace estimation (SE) are popular data analysis tools and used in a wide range of applications. The main interest in PCA/SE is for dimensionality reduction and low-rank approximation purposes. The emergence of big data streams have led to several essential issues for performing PCA/SE. Among them are (i) the size of such data streams increases over time, (ii) the underlying models may be time-dependent, and iii) problem of dealing with the uncertainty and incompleteness in data. A robust variant of PCA/SE for such data streams, namely robust online PCA or robust subspace tracking (RST), has been introduced as a good alternative. The main goal of this paper is to provide a brief survey on recent RST algorithms in signal processing. Particularly, we begin this survey by introducing the basic ideas of the RST problem. Then, different aspects of RST are reviewed with respect to different kinds of non-Gaussian noises and sparse constraints. Our own contributions on this topic are also highlighted.Trung Thanh Leletrungthanhtbt@gmail.comViet Dung Nguyennvdung1987@gmail.comLinh Trung Nguyenlinhtrung@vnu.edu.vnKarim Abed-Meraimkarim.abed-meraim@univ-orleans.fr2021-06-28T02:28:49Z2021-06-28T02:28:49Zhttp://eprints.uet.vnu.edu.vn/eprints/id/eprint/4525This item is in the repository with the URL: http://eprints.uet.vnu.edu.vn/eprints/id/eprint/45252021-06-28T02:28:49ZEfficiently compressing 3D medical images for teleinterventions via CNNs and anisotropic diffusionHa Manh LuuTheo WalsumDaniel FranklinPhuong Cam PhamLuu Dang VuAdriaan MoelkerMarius StaringXiem VanHoangWiro NiessenNguyen Linh Trung2021-06-18T10:55:39Z2021-06-18T10:55:39Zhttp://eprints.uet.vnu.edu.vn/eprints/id/eprint/4454This item is in the repository with the URL: http://eprints.uet.vnu.edu.vn/eprints/id/eprint/44542021-06-18T10:55:39ZPerformance lower bounds of blind system identification techniques in the presence of channel order estimation errorIn this paper, we derive two performance lower bounds
for blind system identification in the presence of channel order
estimation error. The first bound deals with models where
both the channel and unknown symbols are deterministic, and
obtained via the constrained misspecified Cramer-Rao bound
(MCRB). When transmitted symbols are unknown random variables i.i.d. drawn from a stochastic Gaussian process, variance
of any unbiased estimators is always higher than the second
MCRB bound. Both proposed MCRB bounds reduce to the
classical Cramer-Rao bounds when the channel order is known
or accurately estimated. Besides, the stochastic MCRB is lower than the deterministic bound, especially at high SNRsTrung Thanh Leletrungthanhtbt@gmail.comAbed Meraim Karimkarim.abed-meraim@univ-orleans.frLinh Trung Nguyenlinhtrung@vnu.edu.vn2021-06-18T10:55:27Z2021-06-18T10:55:27Zhttp://eprints.uet.vnu.edu.vn/eprints/id/eprint/4453This item is in the repository with the URL: http://eprints.uet.vnu.edu.vn/eprints/id/eprint/44532021-06-18T10:55:27ZA Fast Randomized Adaptive CP Decomposition For Streaming TensorsIn this paper, we introduce a fast adaptive algorithm for CAN- DECOMP/PARAFAC decomposition of streaming three-way tensors using randomized sketching techniques. By leveraging randomized least-squares regression and approximating matrix multiplication, we propose an efficient first-order estimator to minimize an exponentially weighted recursive least- squares cost function. Our algorithm is fast, requiring a low computational complexity and memory storage. Experiments indicate that the proposed algorithm is capable of adaptive tensor decomposition with a competitive performance evaluation on both synthetic and real data.Trung Thanh Leletrungthanhtbt@gmail.comAbed Meraim Karimkarim.abed-meraim@univ-orleans.frLinh Trung Nguyenlinhtrung@vnu.edu.vnHafiane Adel2021-05-31T11:05:29Z2021-05-31T11:05:29Zhttp://eprints.uet.vnu.edu.vn/eprints/id/eprint/4445This item is in the repository with the URL: http://eprints.uet.vnu.edu.vn/eprints/id/eprint/44452021-05-31T11:05:29ZA Hypercuboid-Based Machine Learning Algorithm for Malware ClassificationMalware attacks have been among the most serious threats to cyber security in the last decade. Anti-malware software can help safeguard information systems and minimize their exposure to the malware. Most of anti-malware programs detect malware instances based on signature or pattern matching. Data mining and machine learning techniques can be used to automatically detect models and patterns behind different types of malware variants. However, traditional machine-based learning techniques such as SVM, decision trees and naive Bayes seem to be only suitable for detecting malicious code, not effective enough for complex problems such as classification. In this article, we propose a new prototype extraction method for non-traditional prototype-based machine learning classification. The prototypes are extracted using hypercuboids. Each hypercuboid covers all training data points of a malware family. Then we choose the data points nearest to the hyperplanes as the prototypes. Malware samples will be classified based on the distances to the prototypes. Experiments results show that our proposition leads to F1 score of 96.5% for classification of known malware and 97.7% for classification of unknown malware, both better than the original prototype-based classification method.Thi Thu Trang Nguyentrangngtt@vnu.edu.vnDai Tho Nguyennguyendaitho@vnu.edu.vnDuy Loi Vuvuduyloi55@gmail.com2021-05-31T10:58:32Z2021-05-31T10:58:32Zhttp://eprints.uet.vnu.edu.vn/eprints/id/eprint/4437This item is in the repository with the URL: http://eprints.uet.vnu.edu.vn/eprints/id/eprint/44372021-05-31T10:58:32ZHierarchical Convolutional Neural Network with Feature Preservation and Autotuned Thresholding for Crack DetectionDrone imagery is increasingly used in automated inspection for infrastructure surface defects, especially in hazardous or unreachable environments. In machine vision, the key to crack detection rests with robust and accurate algorithms for image processing. To this end, this paper proposes a deep learning approach using hierarchical convolutional neural networks with feature preservation (HCNNFP) and an intercontrast iterative thresholding algorithm for image binarization. First, a set of branch networks is proposed, wherein the output of previous convolutional blocks is half-sizedly concatenated to the current ones to reduce the obscuration in the down-sampling stage taking into account the overall information loss. Next, to extract the feature map generated from the enhanced HCNN, a binary contrast-based autotuned thresholding (CBAT) approach is developed at the post-processing step, where patterns of interest are clustered within the probability map of the identified features. The proposed technique is then applied to identify surface cracks on the surface of roads, bridges or pavements. An extensive comparison with existing techniques is conducted on various datasets and subject to a number of evaluation criteria including the average F-measure (AFβ) introduced here for dynamic quantification of the performance. Experiments on crack images, including those captured by unmanned aerial vehicles inspecting a monorail bridge. The proposed technique outperforms the existing methods on various tested datasets especially for GAPs dataset with an increase of about 1.4% in terms of AFβ while the mean percentage error drops by 2.2%. Such performance demonstrates the merits of the proposed HCNNFP architecture for surface defect inspection.Qiuchen ZhuQiuchen.Zhu@student.uts.edu.auTran Hiep Dinhtranhiep.dinh@vnu.edu.vnManh Duong Phungduongpm@vnu.edu.vnHa Quangquang.ha@uts.edu.au2021-05-31T10:58:19Z2021-05-31T10:58:19Zhttp://eprints.uet.vnu.edu.vn/eprints/id/eprint/4436This item is in the repository with the URL: http://eprints.uet.vnu.edu.vn/eprints/id/eprint/44362021-05-31T10:58:19ZSafety-enhanced UAV Path Planning with Spherical Vector-based Particle Swarm OptimizationThis paper presents a new algorithm named spherical vector-based particle swarm optimization (SPSO) to deal with the problem of path planning for unmanned aerial vehicles (UAVs) in complicated environments subjected to multiple threats. A cost function is first formulated to convert the path planning into an optimization problem that incorporates requirements and constraints for the feasible and safe operation of the UAV. SPSO is then used to find the optimal path that minimizes the cost function by efficiently searching the configuration space of the UAV via the correspondence between the particle position and the speed, turn angle and climb/dive angle of the UAV. To evaluate the performance of SPSO, eight benchmarking scenarios have been generated from real digital elevation model maps. The results show that the proposed SPSO outperforms not only other particle swarm optimization (PSO) variants including the classic PSO, phase angle-encoded PSO and quantum-behave PSO but also other state-of-the-art metaheuristic optimization algorithms including the genetic algorithm (GA), artificial bee colony (ABC), and differential evolution (DE) in most scenarios. In addition, experiments have been conducted to demonstrate the validity of the generated paths for real UAV operations. Source code of the algorithm can be found at https://github.com/duongpm/SPSO.Manh Duong Phungduongpm@vnu.edu.vnHa Quangquang.ha@uts.edu.au2021-05-31T10:58:07Z2021-05-31T10:58:07Zhttp://eprints.uet.vnu.edu.vn/eprints/id/eprint/4435This item is in the repository with the URL: http://eprints.uet.vnu.edu.vn/eprints/id/eprint/44352021-05-31T10:58:07ZRobust Subspace Tracking with Missing Data and Outliers: Novel Algorithm with Convergence GuaranteeIn this paper, we propose a novel algorithm, namely
PETRELS-ADMM, to deal with subspace tracking in the presence of outliers and missing data. The proposed approach consists of two main stages: outlier rejection and subspace estimation. In the first stage, alternating direction method of multipliers (ADMM) is effectively exploited to detect outliers affecting the observed data. In the second stage, we propose an improved version of the parallel estimation and tracking by recursive least squares (PETRELS) algorithm to update the underlying subspace in the missing data context. We then present a theoretical convergence analysis of PETRELS-ADMM which shows that it generates a sequence of subspace solutions converging to the optimum of its batch counterpart. The effectiveness of the proposed algorithm, as compared to state-of-the-art algorithms, is illustrated on both simulated and real data.Trung Thanh Leletrungthanhtbt@gmail.comViet Dung Nguyennvdung@vnu.edu.vnLinh Trung Nguyenlinhtrung@vnu.edu.vnKarim Abed-Meraimkarim.abed-meraim@univ-orleans.fr2020-07-10T05:41:39Z2020-07-10T05:41:46Zhttp://eprints.uet.vnu.edu.vn/eprints/id/eprint/4003This item is in the repository with the URL: http://eprints.uet.vnu.edu.vn/eprints/id/eprint/40032020-07-10T05:41:39ZAdaptive Algorithms for Tracking Tensor-Train Decomposition of Streaming TensorsTensor-train (TT) decomposition has been an efﬁcient tool to ﬁnd low order approximation of large-scale, high-order tensors. Existing TT decomposition algorithms are either of high computational complexity or operating in batch-mode, hence quite inefﬁcient for (near) real-time processing. In this paper, we propose a novel adaptive algorithm for TT decomposition of streaming tensors whose slices are serially acquired over time. By leveraging the alternating minimization framework, our estimator minimizes an exponentially weighted least-squares cost function in an efﬁcient way. The proposed method can yield an estimation accuracy very close to the error bound. Numerical experiments show that the proposed algorithm is capable of adaptive TT decomposition with a competitive performance evaluation on both synthetic and real data.Thanh Le Trungletrungthanhtbt@gmail.comAbed-Meraim Karimkarim.abed-meraim@univ-orleans.frTrung Nguyen Linhlinhtrung@vnu.edu.vnBoyer Remyremy.boyer@univ-lille.fr