https://hdujs.edu.vn/index.php/hdujs <p>Hong Duc University Journal of Science is the press agency established under the Operating Licence No.14/MIC-OL, dated on January 1st, 2009 and No.125/MIC-OL reissue dated on April 10th, 2014 by the Ministry of Information and Communications. Since 2014, Hong Duc University Journal of Science has been allowed to increase the number periodically to 6 volumes a year in both English and Vietnamese under the International Standard Serial Number: ISSN 1859 - 2759 (for Vietnamesse issue), and ISSN 2588 - 1523 (for English issue) approved by National Agency for Science and Technology Information. Hong Duc University Journal of Science in English is the press reflecting educational and training activities; publishing works, scientific studies of staff, faculties, students, scientists inside and outside the school; propagating and disseminating the policy guidelines and policies of the Party and the State on education and training; introducing, exchanging research results, the application of science and technology in the country and internationally. Editorial Board would like to receive the contribution of numerous faculty members, research scholars, scientists inside and outside the school to Hong Duc University Journal of Science so that we could bring readers the better results, useful information and scientific value.</p> <p>EDITORIAL BOARD</p> Tạp chí Khoa học Trường Đại học Hồng Đức vi-VN Hong Duc University Journal of Science 3030-4768 https://hdujs.edu.vn/index.php/hdujs/article/view/828 2D convolutional neural networks are the key component in partial video copy detection systems. They play a crucial role in video retrieval and matching tasks within a large database. However, the performance characteristics of these feature extraction methods have been little discussed in the literature. This paper presents two key contributions. First, we conduct the experiments on a large-scale dataset to demonstrate the generalization capability and clarify the performance characteristics of popular neural networks. Next, we propose a time-series model approach to highlight the advantages and limitations of image features extracted from neural networks in the partial video copy detection problem. Van Hao Le Dinh Nghiep Le Van Cuong Nguyen Copyright (c) 2025 Hong Duc University Journal of Science 32 39 10.70117/hdujs.79.09.2025.828 https://hdujs.edu.vn/index.php/hdujs/article/view/748 Abstract In this work, an acrylic polymer dispersion and fly ash base coating with good technical properties was successfully prepared. The influence of fly ash content on the technical properties (adhesion, alkali resistance, hot and cold cycle durability and washout durability) of the paint film was investigated. The results showed that the technical properties of the paint film reached the optimal values ​​with fly ash content in the range of 10÷25%. At fly ash content > 25%, the adhesion of the coating decreased. TS T Huong Hoang Truc Vy Do Copyright (c) 2025 Hong Duc University Journal of Science 75 81 10.70117/hdujs.79.09.2025.748 https://hdujs.edu.vn/index.php/hdujs/article/view/771 Cadmium (Cd) has great potential for application in the technology of manufacturing super-strong or heat-resistant alloys. The study of the thermodynamic properties of this metal is necessary not only in the scientific field but also in many industries. However, the study of some thermodynamic quantities of Cd such as Debye frequency and temperature, Debye – waller factor (DWF) at high pressure is still limited. In this paper, we use a semi-empirical method to study some thermodynamic properties of Cd metal including Debye frequency and temperature, DWF. The anharmonic Debye model and the Lindemann’s melting law are combined to study the melting temperature of Cd metal. Numerical calculations have been performed for Cd metal up to 50 GPa. The results we obtained are in good agreement with some previous results. This study not only provides additional data on the thermodynamic properties of Cd but also brings many practical applications in scientific and industrial fields. TS. Thi Hong Nguyen Hoang Dat Nguyen Hai Yen Le Ho Ba Hung Trinh Thi Huyen Trinh Cao Nam Bui Thi Hoang Tran Copyright (c) 2025 Hong Duc University Journal of Science 48 57 10.70117/hdujs.79.09.2025.771 https://hdujs.edu.vn/index.php/hdujs/article/view/836 In this paper, we first state the prolongation problem of ring extensions and show that its data (the triple ) determines a regular E-system. Each extension of this regular E-system is a prolongation of the extension E. Next, we prove the result on the classification of extensions of regular E-systems. Then, we obtain the result on the prolongation problem of a ring extension. Thi Cuc Pham Copyright (c) 2025 Hong Duc University Journal of Science 13 23 10.70117/hdujs.79.09.2025.836 https://hdujs.edu.vn/index.php/hdujs/article/view/923 With the development of industrial sectors, the demand for steel production is increasing, leading to a large amount of blast furnace slag being released. However, the use of this industrial waste in producing innovative concrete is still limited in Vietnam. The objective of this study is to use ground granulated blast furnace slag (GGBFS) to partially replace cement in the production of reactive powder concrete. The control mixture was designed with 75% cement and 25% silica fume as binder materials. Other studied mixtures were formed by replacing 15-60% cement with GGBFS. Test results indicated that the presence of GGBFS contributes to enhancing the concrete workability, reduces the unit weight of fresh concrete, decreases the water absorption, and increases the thermal insulation. Although the ultrasonic pulse velocity of concrete reduced with increasing GGBFS content, all concrete mixtures in this study had a good quality with compressive strength of higher than 75 MPa, ultrasonic pulse velocity of above 4200 m/s, and the water absorption of lower than 3.42%. Based on the experimental program in this study, the optimal GGBFS content used to replace cement was suggested as 15-30%. TS Mai Thi Hong Mai TS Van Dung Nguyen PGS Si Huy Ngo ThS Thi Ha Phuong Trinh Copyright (c) 2025 Hong Duc University Journal of Science 40 47 10.70117/hdujs.79.09.2025.923 https://hdujs.edu.vn/index.php/hdujs/article/view/937 Noise is a concept that is constantly present in our lives, and the study of noise treatment has become an important practical issue. In this article, we will discuss a noise treatment method through wavelet analysis and synthesis algorithms combined with ANC active noise control technology. Active noise control (ANC), also known as noise cancellation (NC), or active noise reduction (ANR), is a method of reducing unwanted sound by adding a second, specially designed sound to cancel out the first sound. ANC uses special equipment to listen to ambient noise and then generate a second audio signal that has all the characteristics but is in the opposite phase to those noises. By the principle of interference, when the two opposite signals are combined, the noise will be canceled. Wavelets are a very suitable tool for us to generate this second signal. Wavelets are wave-like oscillations whose amplitude starts at zero, increases or decreases, and then returns to zero one or more times. Wavelets are colloquially known as "small waves". Wavelets are classified based on the number and direction of their pulses. Wavelets have specific characteristic properties that are very useful for signal processing. Thị Thanh Thủy Trần Copyright (c) 2025 Hong Duc University Journal of Science 82 89 10.70117/hdujs.79.09.2025.937 https://hdujs.edu.vn/index.php/hdujs/article/view/958 In this article, we study the existence and uniqueness of Leray-Hopf solutions and non-negative Leray-Hopf solutions of dyadic models for the 3D generalized Navier-Stokes equations. Our results extend and improve some previous results. Tran Tinh Le Thi Van Pham Chau Giang Le Cam Anh Luu Thi Quy Pham Trinh Nhu Quynh Le Copyright (c) 2025 Hong Duc University Journal of Science 90 102 10.70117/hdujs.79.09.2025.958 https://hdujs.edu.vn/index.php/hdujs/article/view/965 This study investigates the use of spent coffee grounds to replace a part of natural river sand in producing mortars. All mortar mixtures were designed with spread diameters from 155 mm to 185 mm, using cement and lime powder as binder materials, and using spent coffee grounds with sand replacement levels of 0%, 5%, 10%, 15%, and 20% in comparison with binder materials. Test results indicated that unit weight, flexural strength, compressive strength, ultrasonic pulse velocity, and thermal conductivity of mortars decreased with increasing spent coffee grounds amount. However, all spent coffee grounds mortars had a satisfactory quality with compressive strength of above 15 MPa, and ultrasonic pulse velocity of higher than 3000 m/s. This study also indicated that it is possible to use spent coffee grounds to replace natural sand with 20% in comparison with binder materials in producing mortars. Thi Thanh Tam Le Mai Thi Hong Mai Le Sy Chinh Le Sy Chinh Nguyen Van Lam Nguyen Van Lam Copyright (c) 2025 Hong Duc University Journal of Science 65 74 10.70117/hdujs.79.09.2025.965 https://hdujs.edu.vn/index.php/hdujs/article/view/977 This paper presents a semi-automated image labeling and annotation software system that leverages artificial intelligence to support the creation of training datasets for computer vision models. The system integrates the YOLOv8 model for automatic object detection, allowing users to adjust bounding boxes, write and edit image annotations, retrieve data, and download datasets. Additionally, the software offers user access control, data category management, batch image uploads, and flexible image-annotation querying. Featuring a user-friendly, cross-platform interface, the system is easy to deploy in educational and research environments. Experimental results demonstrate that it significantly reduces labeling time while maintaining high accuracy. The proposed solution provides an effective, practical approach to semi-automating training data preparation and also paves the way for future integration of automatic caption generation models. Cong Hoang Anh TS Dinh Cong Nguyen Hoang Long Nguyen The Anh Pham Sam Le Van Copyright (c) 2025 Hong Duc University Journal of Science 5 12 10.70117/hdujs.79.09.2025.977 https://hdujs.edu.vn/index.php/hdujs/article/view/964 Bài báo đưa ra một chứng minh chi tiết công thức phân tích bất khả quy của các iđêan trong vành chính, qua đó xác định mối liện hệ giữa phân tích đa thức thành tích các đa thức bất khả quy với phân tích iđêan thành giao của các iđêan bất khả quy trong vành đa thức hệ số thực và hệ số phức. Dr Xuan Dung Le Thi Hai Yen Nguyen Copyright (c) 2025 Hong Duc University Journal of Science 24 31 10.70117/hdujs.79.09.2025.964 https://hdujs.edu.vn/index.php/hdujs/article/view/953 Nghiên cứu này giới thiệu một phương pháp cải tiến trong đánh giá phát âm tự động (Automatic Pronunciation Assessment - APA) thông qua cơ chế điều chuẩn tương phản thứ bậc. Phương pháp này tập trung vào việc bảo tồn đặc trưng âm vị trong biểu diễn ngữ âm bằng cách kết hợp ba hàm mất mát: tương phản, âm vị và thứ bậc. Mô hình phân tầng gồm ba cấp độ (âm vị, từ và câu) được triển khai nhằm tối ưu hóa khả năng phát hiện và định lượng lỗi phát âm. Kết quả thực nghiệm trên bộ dữ liệu tiếng Anh của người học tiếng Việt cho thấy phương pháp đề xuất cải thiện đáng kể độ chính xác của hệ thống đánh giá phát âm. Phương pháp nghiên cứu mở ra tiềm năng ứng dụng trong các hệ thống học ngôn ngữ thông minh. Thi Bich Nhat Nguyen Copyright (c) 2025 Hong Duc University Journal of Science 58 64 10.70117/hdujs.79.09.2025.953