Global Journal of Computer Science and Technology (GJCST)
Not a member yet
1830 research outputs found
Sort by
Design and Development of an Autonomous Car using Object Detection with YOLOv4
Full text linkFuture cars are anticipated to be driverless point-to-point transportation services capable of avoiding fatalities To achieve this goal auto-manufacturers have been investing to realize the potential autonomous driving In this regard we present a self-driving model car capable of autonomous driving using object-detection as a primary means of steering on a track made of colored cones This paper goes through the process of fabricating a model vehicle from its embedded hardware platform to the end-to-end ML pipeline necessary for automated data acquisition and model-training thereby allowing a Deep Learning model to derive input from the hardware platform to control the car s movements This guides the car autonomously and adapts well to real-time tracks without manual feature-extraction This paper presents a Computer Vision model that learns from video data and involves Image Processing Augmentation Behavioral Cloning and a Convolutional Neural Network model The Darknet architecture is used to detect objects through a video segment and convert it into a 3D navigable path Finally the paper touches upon the conclusion results and scope of future improvement in the technique use
Solution of Integral Equation using Second and Third Order B-Spline Wavelets
Full text linkIt was proven that semi-orthogonal wavelets approximate the solution of integral equation very finely over the orthogonal wavelets Here we used the compactly supported semi-orthogonal B-spline wavelets generated in our paper Compactly Supported B-spline Wavelets with Orthonormal Scaling Functions satisfying the Daubechies conditions to solve the Fredholm integral equation The generated wavelets satisfies all the properties on the bounded interval The method is computationally easy which is illustrated with two examples whose solution closely resembles the exact solution as the order of wavelet increase
Solving the Cubic Monotone 1-in-3 SAT Problem in Polynomial Time
Full text linkThe exact 3-satisfiability problem X3SAT is known to remain NP-complete when restricted to expressions where every variable has exactly three occurrences even in the absence of negated variables Cubic Monotone 1-in-3 SAT Problem The present paper shows that the Cubic Monotone 1-in-3 SAT Problem can be solved in polynomial time and therefore prove that the conjecture P NP hold
Light Deflection in Massive Dyonic Black Holes
Full text linkFollowing Rindler-Ishak method 1 we study the bending of light around general form of dyonic black holes in massive gravity 2 We show that when the Schwarzschild-de Sitter geometry is taken into account does indeed contribute to the bending of ligh
Strengthening Smart Contracts: An AI-Driven Security Exploration
Full text linkSmart contracts are automated agreements in which the conditions between the purchaser and the vendor are encoded directly into lines of code allowing them to execute automatically Smart contracts have emerged as a ground-breaking technology facilitating the decentralized and trustless execution of agreements on blockchain platforms However the widespread adoption of smart contracts exposes them to various security threats leading to substantial financial losses and reputational harm Artificial Intelligence has the capability to aid in the detection and reduction of vulnerabilities thereby enhancing the overall strength and resilience of smart contracts This integration can create highly secure and transparent systems that reduce the risk of fraud corruption and other malicious activities thereby increasing trust and confidence in these systems and improving overall security This research paper delves into the innovative applications of Artificial Intelligence techniques to enhance the security of smart contracts Investigating the potential of AI in detecting vulnerabilities identifying potential attacks and offering automated solutions for safer smart contracts will significantly contribute to the development and flawless execution of this emerging technolog
A Novel Methodology for Generating Demographically Representative Fictional Identities
Full text linkI n an increasingly digitized and data-driven world the capacity to generate synthetic data that can simulate real-world situations is of immense importance It has become particularly relevant in various fields such as data analysis software testing social science simulations and even creative writing These applications often require large sets of data that imitate real- life contexts while ensuring that they are entirely fictional and do not infringe upon individual privacy 9 This paper introduces a novel methodology for creating demographically representative fictional identities specifically designed to reflect the demographic distribution of the United States Creating synthetic identities that match specific demographic distributions presents several benefit
Software System Model Correctness using Graph Theory: A Review
Full text linkThe Unified Modeling Language UML is the de facto standard for object-oriented software model development The UML class diagram plays an essential role in design and specification of software systems The purpose of a class diagram is to display classes with their attributes and methods hierarchy generalization class relationships and associations general aggregation and composition between classes in one mode
Comparison of Prim and Kruskal’s Algorithm
Full text linkThe goal of this research is to compare the performance of the common Prim and the Kruskal of the minimum spanning tree in building up super metric space We suggested using complexity analysis and experimental methods to evaluate these two methods After analysing daily sample data from the Shanghai and Shenzhen 300 indexes from the second half of 2005 to the second half of 2007 the results revealed that when the number of shares is less than 100 the Kruskal algorithm is relatively superior to the Prim algorithm in terms of space complexity however when the number of shares is greater than 100 the Prim algorithm is more superior in terms of time complexity A spanning tree is defined in the glossary as a connected graph with non-negative weights on its edges and the challenge is to identify a maz weight spanning tree Surprisingly the greedy algorithm yields an answer For the problem of finding a min weight spanning tree we propose greedy algorithms based on Prim and Kruskal respectively Graham and Hell provide a history of the issue which began with Czekanowski s work in 1909 The information presented here is based on Rose
Image Compression using Walsh Functions
Full text linkCurrently image transfer and storage require compression to reduce memory usage and to increase transmission speed In this article the hybrid compression algorithm is used for color black and white images It includes the discrete wavelet transform and the Walsh transform are used for quantization The Walsh transform coefficients are quantized and arithmetically encoded The combined output is compressed and can be transmitted over any available network in the shortest time The compressed image is decoded and the original image is decompressed using the inverse conversion operatio
A Novel Frequent Pattern Mining Algorithm for Evaluating Applicability of a Mobile Learning Framework
Full text linkThe applicability of a mobile learning system reflects how it works in an actual situation under diverse conditions In previous studies researches for evaluating applicability in learning systems using data mining approaches are challenging to find The main objective of this study is to evaluate the applicability of the proposed mobile learning framework This framework consists of seven independent variables and their influencing factors Initially 1000 students and teachers were allowed to use the mobile learning system developed based on the proposed mobile learning framework The authors implemented the system using Moodle mobile learning environment and used its transaction log file for evaluation Transactional records that were generated due to various user activities with the facilities integrated into the system were extracted These activities were classified under eight different features i e chat forum quiz assignment book video game and app usage in thousand transactional rows A novel pattern mining algorithm namely Binary Total for Pattern Mining BTPM was developed using the above transactional dataset s binary incidence matrix format to test the system applicability Similarly Apriori frequent itemsets mining and Frequent Pattern FP Growth mining algorithms were applied to the same dataset to predict system applicabilit