Procedia Computer Science最新文献

{"title":"Empowering the Visually Impaired: YOLOv8-based Object Detection in Android Applications","authors":"Shraddha S. More ,&nbsp;Neeta Patil Dr. ,&nbsp;Vivian Brian Lobo ,&nbsp;Nikhil Shet ,&nbsp;Dhruv Goswami ,&nbsp;Praful Rane","doi":"10.1016/j.procs.2025.01.005","DOIUrl":"10.1016/j.procs.2025.01.005","url":null,"abstract":"<div><div>This research helps visually impaired individuals enhance their quality of life using object detection technology. Vision is a fundamental sense that lets people interact and navigate through their surroundings. Earlier, object detection was implemented with methods like Sliding Window, and Haar Cascades which lacked the accuracy and efficiency that are necessary for a real-time environment. This study proposed an object detection system specifically defined for visually impaired individuals using the YOLOv8 model because YOLO is known for its real-time performance, accurate detection, and classification capabilities, ensuring instant and accurate assistance in real-time environments. Additionally, YOLO’s robustness in detecting objects in various scenarios makes it a reliable solution for aiding visually impaired individuals in navigating their surroundings effectively. The novelty of the study lies in converting the YOLOv8 model into TensorFlow Lite, making it easier to deploy on Android platforms and widening its accessibility to a larger user population. The application of this object detection system extends to various aspects of everyday life, promoting inclusivity and accessibility for visually impaired individuals and bridging the gap between cutting-edge research and practical assistive technologies with the help of support on edge devices and auditory feedback. The performance metrics are evaluated on multiple models to understand the performance of the proposed model on different training parameters such as model size and input image resolution. The proposed model with an input image size of 640X640 resolution and Nano-sized model achieved mAP50 and mAP50-95 of 79.9% and 57.1% respectively with inference time of 350ms on average. For the same resolution the proposed model is trained on Medium-sized model and attained mAP50 and mAP50-95 as 80.5% and 60.9% respectively with inference time of 1800ms average. For an image size of 256X256 resolution and Nano-sized model the proposed model achieved mAP50 and mAP50-95 of 74.4% and 52.9% respectively with inference time of 55ms on average. From results it is observed that the model accuracy increases as the model size and input resolution is high but it increases the inference time. Whereas the model with less inference time lower input image resolution gives most efficient results but also reduces accuracy.</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"252 ","pages":"Pages 457-469"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143377149","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A novel stable feature selection algorithm for machine learning based intrusion detection system","authors":"Sowmya T ,&nbsp;Mary Anita E A Dr","doi":"10.1016/j.procs.2025.01.034","DOIUrl":"10.1016/j.procs.2025.01.034","url":null,"abstract":"<div><div>The advent of new technologies like artificial intelligence, and big data has influenced many cyber attackers to launch their attacks on the network. Hence researchers have already proposed Intrusion Detection Systems by incorporating machine learning as well. Building an effective IDS is still a challenging task because of low accuracy. Managing high dimensional data is another major problem that occurs in IDS. Hence in this paper, an efficient Machine Learning based Intrusion Detection System is developed by means of a novel stable feature selection strategy called IV-RFE. The proposed methodology aims to select only the relevant features that contribute to the attack, which is purely based on relative variance, and weight factor in combination with RFE. This methodology increases the performance in terms of accuracy and maintains a stable set of features. Previous studies only focussed on the feature selection strategy and their performance. The feature stability also has to be considered which is an equally important metric, especially in the field of Intrusion Detection Systems. Hence in the current study, an efficient ML based IDS is proposed which selects only the relevant and stable features. Experimental results also revealed that the proposed IV-RFE outperformed well for three attacks with respect to accuracy and stability metrics also. The results show that stability is also an important indicator in selecting the features in the field of Intrusion Detection Systems.</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"252 ","pages":"Pages 738-747"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143376856","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"VBDPA Multi-Criteria Task Scheduling Algorithm in Container Based Cloud Computing Environment","authors":"Himanshukamal Verma ,&nbsp;Vivek Shrivastava","doi":"10.1016/j.procs.2025.01.020","DOIUrl":"10.1016/j.procs.2025.01.020","url":null,"abstract":"<div><div>In today’s dynamic technology world, cloud computing environments can accomplish effective outcomes using container-based solutions. Cloud resource management and allocation have altered as an outcome of containerized cloud computing systems. It provides scalable and flexible solutions for effective resource utilization and various factors regarding businesses. Task scheduling is very challenging in cloud systems that use containerization because there are many competing factors to consider. This study suggests VBDPA, a unique multicriteria-based task-scheduling algorithm for containerized cloud environments. The algorithm effectively addresses the complex and dynamic nature of cloud computing by utilizing the multi-criteria decision-making method known as VIKOR (VIekriterijumskoKOmpromisnoRangiranje). VBDPA seeks to improve overall system performance, reduce task completion times, maximize resource utilization, and provide better revenue for cloud providers. An analysis is carried out to compare VBDPA’s effectiveness with the current Promethee-II based PBDPA algorithm. The results of the experiments in the Cloudsim simulator show that VBDPA performs better than PBDPA in terms of makespan, revenue generation, and resource utilization. The results highlight VBDPA’s potential as a viable option for effective task scheduling in containerized cloud environments</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"252 ","pages":"Pages 603-612"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143376900","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Yawning Detection for Cognitive Distraction in Drivers Using AlexNet: A Deep Learning Approach","authors":"Aakash Kumar ,&nbsp;Kavipriya G ,&nbsp;Amutha S ,&nbsp;Dhanush R","doi":"10.1016/j.procs.2024.12.007","DOIUrl":"10.1016/j.procs.2024.12.007","url":null,"abstract":"<div><div>The usage of cars has surged dramatically, with individuals preferring personal or rental vehicles for daily commuting and vacations over public transportation due to comfort and convenience. However, this preference has led to an increased risk of drowsy driving, which often results in severe accidents and fatalities, contributing to a rise in mortality rates. To address this recognizing yawning in drivers is a critical safety measure, which poses a unique challenge because of the action’s subtlety and variability. This paper proposes a yawning recognition technique utilizing AlexNet, a convolutional neural network recognized for its efficacy in image classification. The approach integrates AlexNet with image pre-processing techniques to enhance the recognition of facial gestures from static, frontal profile-view color images. It also employs various neuron-wise and layer-wise visualization methods on an AlexNet model trained with a publicly available dataset. The results underscore the effectiveness of neural networks in accurately detecting the distinctions of yawning, showcasing the potential of deep learning in improving driver safety.</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"252 ","pages":"Pages 63-72"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143376724","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Application of LSTM-GRU combined model to calculate the reliability of software systems","authors":"Tamilla A. Bayramova ,&nbsp;Tofig H. Kazimov","doi":"10.1016/j.procs.2024.12.014","DOIUrl":"10.1016/j.procs.2024.12.014","url":null,"abstract":"<div><div>This study investigated the effectiveness of deep learning models in assessing the reliability of software systems and the application of recurrent neural network algorithms in reliability prediction. A hybrid model consisting of a combination of LSTM and GRU models is proposed to predict the reliability of software systems. Along with historical data collected during testing and implementation, several environmental factors covering the software life cycle and affecting its reliability, as well as the complexity of the software code, are taken as input. Based on these data, a new method for expert assessment of software reliability is proposed, and the calculated expert scores are taken as output. The proposed model is trained based on these values. This is a comprehensive approach to assessing the reliability of software systems.</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"252 ","pages":"Pages 127-135"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143376731","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Performance Optimization in Blockchain Networks for Healthcare Systems Using Adaptive Sharding","authors":"Sharad Katkol ,&nbsp;Praveen M Dhulavvagol ,&nbsp;S G Totad","doi":"10.1016/j.procs.2025.01.048","DOIUrl":"10.1016/j.procs.2025.01.048","url":null,"abstract":"<div><div>Blockchain technology has the potential to transform healthcare data management by enhancing security, transparency, and data integrity. However, scalability, latency, and privacy concerns have limited its application in high-volume, sensitive environments such as healthcare. This paper introduces a blockchain architecture that addresses these challenges through adaptive sharding and rule-based data partitioning. The adaptive sharding algorithm dynamically adjusts shard configurations in response to real-time network demands, optimizing resource allocation and improving scalability. Meanwhile, rule-based data partitioning organizes transactions across shards based on specific attributes, such as transaction type or geographic region, to minimize cross-shard communication and improving processing efficiency. Together, these methods increase transaction throughput by 34% and reduce latency by 9% compared to traditional approaches. Additionally, the system incorporates Byzantine Fault Tolerance (BFT) consensus to strengthen security, along with zero-knowledge proofs and homomorphic encryption to protect sensitive patient data during transaction verification. This architecture provides a comprehensive, scalable, and secure blockchain solution tailored to the unique needs of healthcare data management, addressing critical limitations while maintaining privacy and data integrity.</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"252 ","pages":"Pages 873-882"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143376769","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"WOCP: Controller Placement using Whale Optimization in SDN-WAN","authors":"Vidya Sagar Thalapala ,&nbsp;Koppala Guravaiah","doi":"10.1016/j.procs.2025.01.049","DOIUrl":"10.1016/j.procs.2025.01.049","url":null,"abstract":"<div><div>The established SDN multi-controller architecture faces the challenge of locating controllers to boost the scalability, reliability, and security of the network. Reliability is one of the important metric to handle when a failure event occurs in a network. In this paper, a Whale Optimization technique is used for controller placement and proposed Whale Optimization based Controller Placement (WOCP), that places the controllers in an optimized and efficient manner. The adequate placement of controllers minimizes the overall average latency with a high reliability index. The WOCP method resemblances with Minimizing Latency for Controller Placement (MLCP), resulting in an 18% reduction in latency in the Geant network topology.</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"252 ","pages":"Pages 883-892"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143376770","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Large Language Model based Educational Virtual Assistant using RAG Framework","authors":"Umair Hasan Khan ,&nbsp;Muneeb Hasan Khan ,&nbsp;Rashid Ali","doi":"10.1016/j.procs.2025.01.051","DOIUrl":"10.1016/j.procs.2025.01.051","url":null,"abstract":"<div><div>In recent times significant advancement have been made in text based chatbots or virtual assistants. This research study presents a latest and transformational approach to provide information support in university through the development of educational virtual assistant based on large language models and retrieval augmented generation framework. Moving forward, the technological advancement have increased the complexity of university systems, so finding innovative solution for students to access information on various topics such as admission processes, course selection, and campus facilities etc. is important. The proposed virtual assistant leverages transformer architecture based large language models specifically Meta-llama/Llama-2-7b-chat-hf and Mistralai/Mistral-7B-Instruct-v0.2 to generate human like text for student inquires. The system design includes a data retrieval process from university websites followed by data pre-processing. Building on retrieval augmented generation framework, the virtual assistant retrieves the most relevant data from a comprehensive university knowledge base, ensuring responses stay updated and precise. The virtual assistant was tested on a range of varieties of university related queries, and its response were evaluated using bilingual evaluation understudy score metrics. Experimental results suggest that the retrieval augmented generation-based Llama-2-7b-chat-hf provides a viable solution for addressing the challenge of providing university related information to students. The findings indicate that retrieval augmented generation based large language models hold significant potential for automating administrative support in educational institutions.</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"252 ","pages":"Pages 905-911"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143376772","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Identity Authentication System Network for Internet of Things and Blockchain-enabled Trading Platforms in Renewable Energy Generating Units","authors":"Partha Pratim Bhattacharjee ,&nbsp;Chaitali Koley ,&nbsp;Saibal Chatterjee","doi":"10.1016/j.procs.2025.01.057","DOIUrl":"10.1016/j.procs.2025.01.057","url":null,"abstract":"<div><div>This article presents a Blockchain-based spotless, efficient clean, and green energy trading platform for environmentally friendly energy-generating units. A system that provides a constant information-obtaining framework for exchanging information and surveillance of these renewal energy-generating units and the energy uses of these units, the health of units, and performance. It typically works in a troublesome climate, and its security has forever been challenging. Node identification proof and authentication is a significant security worry for distributed IoT frameworks; blockchain technology with decentralization highlights another era of arrangements. Data/ Information is gathered and pre-processed utilizing the \"Raspberry Pi” microcontroller Board linked alongside field sensors framing an \"Edge Node (EN)\", which comprises an inbuilt Wi-Fi unit, a generated terminal voltage sensor, generated as well as load current sensors, an external real-time clock circuit, and other environmental surveillance sensor units for fundamental support of the generating unit. These IoT-based \"Wireless Sensor Network (WSN)\" nodes are separated into \"Edge Node (EN)\", \"Fog Node (FN)\", and \"Cloudlet Server (CS)\", as indicated by their capacities and work conveyances which are framed into various leveled network. In this model, nodes’ identity and mutual authentication across multiple communication scenarios are realized through, a private Blockchain developed among the terminal nodes, an EN, forming a complete Blockchain-based wireless sensor network (BWSN).</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"252 ","pages":"Pages 964-974"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143376778","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Kalman filtering based preprocessing for secure key generation","authors":"Tapesh Sarsodia ,&nbsp;Uma Rathore Bhatt ,&nbsp;Raksha Upadhyay ,&nbsp;Vijay Bhat","doi":"10.1016/j.procs.2024.12.042","DOIUrl":"10.1016/j.procs.2024.12.042","url":null,"abstract":"<div><div>The global market of IoT devices is increasing rapidly. Examples of IoT like networks include smart cities, industrial enterprises, agriculture, home automation, healthcare etc. IoT offers efficient resource utilization, enhanced data collection, minimum human efforts etc. although it is constrained by many challenges such as security, privacy, limited interoperability, complexity and integration challenges. Among all, security and privacy are paramount and require efficient techniques with low power and minimum computer complexity as IoT is a power-constrained network. Traditional encryption methods fail to meet these limitations, so physical layer key generation (PLKG) using Received Signal Strength Indicator (RSSI) preprocessing, is a promising approach for securing such wireless networks. In this paper, the use of Kalman filtering for RSSI preprocessing in secure key generation at the physical layer is presented and compared its performance with the existing Principal Component Analysis (PCA) based preprocessing technique. The performance of the proposed approach is evaluated on three fading channels namely Rician, Rayleigh, and Nakagami to highlight its effectiveness in different environments. The results show that the Kalman filtering is significantly better than PCA in terms of Bit Disagreement Rate (BDR), Spearmen rank Correlation Coefficient (SCC) and Entropy, thus providing stronger security guarantees and more reliable key generation. This makes Kalman filtering a potential solution for PLKG in IoT environments, focusing on computing performance and high security.</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"252 ","pages":"Pages 414-423"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143377144","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}