Scopus İndeksli Yayınlar Koleksiyonu
Permanent URI for this collectionhttps://hdl.handle.net/20.500.12416/8651
Browse
5 results
Search Results
Conference Object Multi-Agent OSINT Architecture with Graph RAG Integration and Hierarchical Bloom-Filter Deduplication(Institute of Electrical and Electronics Engineers Inc., 2025) Arslan, Serdar; Yurtalan, GokhanConference Object Enhancing File Security with an Optimized Auto-Classification Framework Based on Learning Models(Institute of Electrical and Electronics Engineers Inc., 2025) Açikgöz, Zeliha; Arslan, Recep Sinan; Arslan, SerdarArticle Enhancing Content-Based Retrieval Through an End-to-End Approach Utilizing Deep Learning and Multidimensional Indexing(Springer London Ltd, 2025) Uzel, Omer; Arslan, SerdarRecent advancements in technology, coupled with reductions in hardware and software costs, have propelled visual search applications into the spotlight, making them both popular and indispensable. Consequently, the rapid and precise retrieval of images from vast databases through image queries has become a critical task. We introduce a novel end-to-end retrieval architecture that significantly enhances retrieval performance when compared to a baseline system that conducts database searches at the video frame level. Leveraging a pre-trained convolutional neural network model, we employ unsupervised image retrieval processes to extract and store low-level features for efficient indexing. To facilitate swift and effective access, we implement a tree-based indexing structure known as VP-Tree. This structure utilizes the extracted low-level features. To make these features compatible with our system, we employ dimension reduction techniques to represent them in a lower-dimensional space. Our experiments, conducted on three benchmark datasets, demonstrate that VP-Tree consistently outperforms k-nearest neighbor (KNN) search in terms of retrieval accuracy and efficiency. Specifically, for image data set, VP-Tree achieves a precision of 56.3903, an F1-score of 68.703, and an area under the curve (AUC) of 93.518719, all slightly surpassing KNN. Similarly, for news video data set, VP-Tree attains a precision of 38.704011, an F1-score of 55.029674, and an AUC of 64.6412, again outperforming KNN. For documentary data set, VP-Tree achieves a notable improvement with a precision of 73.511723, an F1-score of 84.734013, and an AUC of 80.981328, demonstrating superior performance over KNN. In addition to accuracy, we evaluated retrieval time across different dataset sizes. While KNN performs slightly faster on smaller datasets, VP-Tree scales significantly better as dataset size increases. For 100,000 images, VP-Tree reduces retrieval time from 79.77 to 54.34 ms, and for 200,000 images, it improves performance from 108.75 to 44.63 ms, confirming its efficiency in large-scale retrieval scenarios. These results highlight VP-Tree as a robust and scalable alternative to traditional KNN-based methods, ensuring both accuracy and efficiency in large-scale image retrieval tasks.Article Citation - Scopus: 6Design of Frameless Gimbal Motor for Uav Applications(International Organization on 'Technical and Physical Problems of Engineering', 2022) Arslan, S.; İskender, İres; Iskender, I.; Arslan, Serdar; Elektrik-Elektronik Mühendisliği; Bilgisayar MühendisliğiRecently, application areas of the Unmanned Aerial Vehicle (UAV) systems have started to expand very rapidly due to the fact that offering more effective, economical, reliable and safe solutions compared to manned air platforms, satellites and/or various ground platforms. However, desire to develop higher performance, resourceful, lighter, small and low powered payload make the gimbal platforms mandatory part of the UAVs in a short time and their role is getting increased day by day. In parallel with the increasing demand for precise stabilization, robustness, lightness and agility in gimbal systems, it has become an important trend to use more-electric (ME) customed systems instead of traditional market products. The electric motors that control the speed and position of the gimbal system are simply referred to as gimbal motors. Related design study focuses on designing direct-drive in-runner frameless gimbal motor with the following features; 8.5 VAC line voltages, 24-slot/28-pole combination, 60 rpm, 80 mN.m. Permanent magnet synchronous motor topology is determined to offer higher torque density, higher precision and fast response required for gimbal platforms. The selecting criteria of dimensions, performance parameters, materials, machine type with rotor structures and motor duty cycle are also explained. The gimbal motor is performed analytically in Ansys RMxprt with parametric assignments, statistically and sensitively tuned in Maxwell 2D and optimized in Maxwell 3D by finite element method (FEM) optimetric convergence approach with magnetostatic and transient solutions to get the final machine shape. This study is currently part of the gimbal system to be produced for medium sized surveillance UAV. Since the gimbal motor has been prototyped, all dimensions given are valid. © 2022, International Organization on 'Technical and Physical Problems of Engineering'. All rights reserved.Article Citation - WoS: 1Citation - Scopus: 2Redefining Osint Software Architecture With System-Centric Architecture Design: a Framework Shaped by Qaw, Add, and Atam(Ieee-inst Electrical Electronics Engineers inc, 2025) Yurtalan, Gokhan; Arslan, SerdarThis study develops a novel software architecture for Open Source Intelligence (OSINT). The primary architectural drivers of the OSINT architecture are identified using the Quality Attribute Workshop (QAW), and an end-to-end OSINT software architecture design is implemented in accordance with Attribute-Driven Design (ADD). The architecture is extensively analyzed with metric evaluations and the Architecture Tradeoff Analysis Method (ATAM), confirming critical quality attributes such as performance, reliability, functional suitability, and security. The design decisions taken within this architectural framework are detailed in the article through module view, component and connector view, and allocation view representations. The proposed architecture uses an on-premise Large Language Model (LLM) to explore the potential for deeper and more reliable information processing capabilities in OSINT analyses and presents a framework that enhances semantic depth and analytical capabilities. The architecture not only amplifies the semantic and analytical capabilities of OSINT systems but also sets a precedent for future architectural endeavors in intelligence systems design. This paper presents a framework that not only meets contemporary needs but also anticipates future demands in the rapidly evolving field of OSINT.