Titus Issac
@karunya.edu
Assistant Professor
Karunya Institute of Technology & Sciences
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
Scopus Publications
Immanuel John Raja, S. V. Evangelin Sonia, C. P. Shirley, and I. Titus
Springer Science and Business Media LLC
C. P. Shirley, J Immanuel John Raja, S. V. Evangelin Sonia, and I. Titus
Springer Science and Business Media LLC
M Udhayakumar, Titus Issac, and J Sebastian Terance
IEEE
The communication between differently challenged persons especially deaf and mute persons to a common man is challenging. The research work proposes to investigate various deep learning models to enhance communication between the deaf and persons who may not know sign language. The work uses computer vision technology for real-time hand gesture identification and natural language processing to produce coherent and understandable words This article investigates the feasibility of applying deep learning techniques like convolutional neural networks, xgboost and random forest to recognize hand gestures in real-time. The models are trained on a dataset of hand gestures and are investigated to identify hand gestures in real-time. The model can recognize a wide range of hand motions, from simple ones like pointing and waving to complex ones involving several fingers. The system creates sentences that are consistent with the hand gesture's intended meaning once it has been identified. Based on established rules and grammar suggestions, the model tries to generate appropriate sentences while taking the context of the conversation into account. The proposed work would be highly beneficial to researchers and target audiences who have trouble expressing themselves in a quick, efficient way to communicate.
Amish Abraham G, Titus Issac, and J Sebastian Terrance
IEEE
This research study aims to enhance the intelligent classroom ecosystem by integrating the Local Binary Pattern Histogram (LBPH) algorithm with human expertise to identify students' faces. Monitoring student attentiveness is crucial for effective learning. To support this, this study uses Mediapipe to analyze students' body posture, providing insights into their engagement levels. Additionally, the proposed framework includes the Eye Aspect Ratio (EAR) Algorithm to detect signs of drowsiness and yawning. By continually refining the proposed framework, this study intends to improve intelligent classroom ecosystems, helping educators to create environments that increase student success.
Praisy Mizpah Julia Bestus, Titus Issac, J. Sebastian Terence, and T. D. Subash
IEEE
In recent years, Unmanned Aerial Vehicles (UAVs) have become increasingly prevalent in various applications, including surveillance, monitoring, and delivery services. Efficiently assigning tasks to UAVs while optimizing multiple objectives such as distance travelled, energy consumption, and task completion time presents a complex optimization problem. To address this challenge, the approach put forward is a Multi-Objective Grey Wolf Algorithm (MOGWA) specifically tailored for solving UAV task assignment problems in a two-dimensional (2D) environment. The MOGWA is an extension of the classic Grey Wolf Algorithm (GWA), which draws inspiration from the social behaviour and hunting strategies of grey wolves. In the proposed MOGWA, UAVs are represented as individual grey wolves, and their movements are guided by principles of hierarchical leadership within a wolf pack. This allows for a balance between exploration and exploitation of the solution space, enabling efficient search for optimal task assignments. Experimental evaluations demonstrate the effectiveness of the MOGWA in finding high-quality solutions for UAV task assignment problems in 2D environments. The proposed algorithm offers a promising framework for addressing real-world challenges in UAV mission planning and optimization, particularly in scenarios where multiple conflicting objectives need to be considered simultaneously.
X Samuel Roy, Titus Issac, and J Sebastian Terrance
IEEE
A travel itinerary is a complex problem that involves multiple objectives and constraints, such as cost, time, transportation modes, and comfort levels. This research study focuses on creating a real-time eco-friendly itinerary generator by integrating machine learning techniques. This research work explores two machine-learning algorithms: Artificial Neural Networks (ANNs) and Multi-Layer Perceptron (MLP). These algorithms are evaluated for their suitability in generating user-friendly and environment-friendly itineraries. Through experimentation, the accuracy and precision of the proposed models are assessed, leading to the identification of an optimal travel planner model. The ultimate goal of this model is to provide a detailed itinerary that considers prominent modes of transportation while addressing the pressing environmental crisis. By prioritizing eco-friendly options, this approach sets itself apart from conventional travel planning tools.
K. Manaswitha, Titus Issac, Salaja Silas, and J Sebastian Terance
IEEE
In recent years, there have been significant advancements in Unmanned Aerial Vehicles (UAVs), leading to their integration into everyday life. UAVs are constrained by limited energy, communication, and localization. The various challenges in UAV research include localization, routing, navigation, and task assignment. Recently, UAV swarms have been employed across a wide range of applications. This study focuses on solving task assignment problems within UAV swarms, considering the unique multi-objective resource constraints of UAVs. The work explores the feasibility of adopting the Red Fox Algorithm to address the task assignment problem in UAV swarms. The investigation examines two primary input parameters: Task Energy Demand and distance. The evaluation includes energy utilization and shortest distance as performance metrics. Notably, the proposed approach outperforms the Particle Swarm Optimization Algorithm (PSO), resulting in a 17.47% lower energy utilization while maintaining a 17.83% reduction in the shortest distance travelled. By leveraging the Red Fox Algorithm, the work aims to enhance the efficiency and effectiveness of task allocation in UAV swarms, contributing to their broader adoption across various applications.
Titus Issac, Salaja Silas, and Elijah Blessing Rajsingh
Springer Science and Business Media LLC
Titus Issac, Salaja Silas, and Elijah Blessing Rajsingh
Elsevier BV
Titus Issac, Salaja Silas, and Elijah Blessing Rajsingh
Springer Singapore
Titus Issac, Salaja Silas, and Elijah Blessing Rajsingh
Springer Science and Business Media LLC
Titus Issac, Salaja Silas, Elijah Blessing Rajsingh, and Sharmila Anand John Francis
Elsevier
T.D. Subash, T.D. Subha, I. Titus, Alsufiyan Nazim, and Eugene Peter
Elsevier BV
Titus Issac, Salaja Silas, and Elijah Blessing Rajsingh
Elsevier
Titus Issac, Salaja Silas, and Elijah Blessing Rajsingh
IEEE
Modern heterogeneous wireless sensor nodes can be used to develop a wide plethora of sophisticated Wireless Sensor Network (WSN) applications. In a WSN, the nodes collaborate with each other to achieve the desired objectives by employing a task assignment algorithm. The majority of the existing WSN task assignment algorithms were designed for a homogeneous environment. However, the current trend of using heterogeneous nodes in WSN application warrants an elaborate investigations on the various factors influencing task assignment in heterogeneous environment. Extensive analysis on decisive factors such as node properties, WSN architecture, WSN application types were exhaustively carried out. Subsequently, a multi-objective based task assignment algorithm using Particle Swarm Optimization (PSO) was proposed. Various case studies on PSO by varying the fitness function and criteria weights were modelled and experimented through simulation to study the feasibility of achieving the desired objectives. The performance metrics such as energy consumption, response time and successful task assignment ratio were analyzed under different cases. Our investigations reveal that multi-objective based PSO outperforms its legacy counterpart in achieving the desired objectives with higher successful task assignment ratio in the heterogeneous environment.
Titus Issac, Salaja Silas, and Elijah Blessing Rajsingh
IEEE
Devices are becoming smarter with the incorporation of sophisticated sensors, micro controllers and communication devices. A smart city integrates its core services using smart devices. The outdoor lighting application is one of the key core services, as it is a high power demand application. With the advent of sophisticated, smart luminaries, the need for an energy efficient, green smart city and adequately well-lit safe smart city has been much awaited. A safe green smart city is achieved by meticulously assigning illumination roles to every individual luminaries via a dynamic role assignment scheme. The proposed dynamic scheme assigns the illuminating task to the luminaries based on the evaluation criteria such as luminaire life span, illumination level, temperature, peak hours and critical zones. The objectives of the work is to achieve a green smart city by reducing power consumption and provide a safer smart city with adequate illumination. The proposed dynamic role assignment has been compared with the conventional lighting methods and static role assignment algorithm (COIRAS). The analyses reveal the proposed dynamic role assignment scheme achieve the objectives for a safe and green smart city.
Titus Issac, Salaja Silas, and Elijah Blessing Rajsingh
Elsevier BV
Titus Issac, Salaja Silas, and Elijah Blessing Rajsingh
Springer Singapore
Salaja Silas, Elijah Blessing Rajsingh, and Titus Issac
IEEE
Recently there has been an evidential growth in E-healthcare services. Every hospital has a variety of similar or dissimilar healthcare services. Selecting the best healthcare service is influenced by many preferences such as doctor's experience, location, feedback on continuity of care, waiting time, cost, hospital facilities, etc. Among the preferences, feedback is more influential. Participant's feedback plays a vital in selection of the best healthcare service and also in improving the quality of the healthcare service and its provider. In this paper, weighted average based subjective feedback system has been proposed, designed and implemented to obtain the feedback from the various personnel related to the e-health care services. Experimental analyses have been conducted to prove that the proposed feedback system is effective.
G. Edwin Prem Kumar, I. Titus, and Sony. I. Thekkekara
Elsevier BV