Mobile users can open program in new tab for better viewing.

Open program in new tab

Day 1 18/12/2020
Room #1

Opening Speech 09:00 - 09:45

Keynote Speech 1 09:45 - 10:45

Prof. Dr. Joan Lu

Tea Break 10:45 - 11:00

Technical Session 1 11:00 - 13:00

11:00 - 11:20
Color-to-Grayscale conversion for images with non-uniform chromatic distribution using Multiple Regression

Color-to-Grayscale conversion methods try to identify weights for various color channels for obtaining a gray-scale image. These weights can be fixed either globally or computed on a localized basis. This paper presents an approach for computing the global weights using localized regions chosen using the assistance of Human Vision System. For a given image, the proposed method aims to maximize the required foreground information, which is normally present in the dominant color channel. The proposed method was tested on DIBCO-2013 dataset and qualitatively evaluated using PSNR, MSE and SSIM. The results obtained have established to be more satisfactory. The experimental results of ours and other color-to-grayscale methods have been tabulated and discussed.
Authors: Paramasivam ME (Sona College of Technology,Salem), Sabeenian RS (Sona College of Technology), Dinesh PM (Sona College of Technology,Salem), Anand R (Sona College of Technology, Salem, India), Eldho Paul (Sona College of Technology),
Hide Authors & Abstract

Show Authors & Abstract
11:20 - 11:45
Reorganizing Virtual Machines as Docker Containers for Efficient Data Centers

Increase in the use of virtual machines in various fields there is a need to enhance the balancing of huge workload in data centres. The traditional way of implementing the cloud usually done with virtual machines. With working of virtualization can be reconsidered with different and enhanced technology such as containers in data centre. Docker containers are gaining popularity due to its features like increase in productivity by reducing the number of re-sources. In this paper, virtual machines are compared with the containers and virtualization techniques are examined so that the user can work with respect to the requirements. Unlike a virtual machine, container does not have another software layer called Hypervisor. Due to these reasons containerized applications have better performance characteristics than virtual machine-based applications. We will be discussing on Benefits of containers over hyper-visor in containers. Nevertheless, containers execute directly in the kernel of the virtual machine. Docker containers use engine of the Docker as an alternative to hypervisor.
Authors: Vasantha Kumari N (presidency university), Arulmurugan Ramu (Presidency university),
Hide Authors & Abstract

Show Authors & Abstract
11:45 - 12:20
Similarity analytics for semantic text using Natural Language Processing

Determining the similarity among the sentences is a predominant task in natural language processing. The semantic determining task is one of the im-portant research area in today‟s applications related to text analytics. The se-mantic of the sentences get varied according to the textual context it is used. In natural language processing, determining the semantic likeness between sen-tences is an important research area. As a result, a lot of research is done in de-termining the semantic likeness in the text. For example, there exists many pos-sible semantics for a word (polysemy), and the synonym of the word; and also these techniques avoid considering the stop words in English which are critical for English phrase/word division, speech investigation, and meaningful com-prehension. Our proposed work utilizes Term_Frequency based Inverse_Document_Frequency model and Glove algorithm based word_embeddings vector for determining the semantic similarity among the terms in the textual contents. Lemmatizer is utilized to reduce the terms to the most possible smallest lemma‟s. The outcomes demonstrate that the proposed methodology is more prominent than the TF-idf score in ranking the terms with respect to the search query terms. The pearson correlation coefficient achieved for the semantic similarity model is 0.875.
Authors: Karthiga M (Bannari Amman Institute of Technology), sountharrajan S (VIT Bhopal University), Suganya E (Anna University Chennai), Bazila Banu A (Bannari Amman Institute of Technology), Sankarananth S (Excel College of Engineering and Technology), Sathish Kumar B (VIT University Chennai),
Hide Authors & Abstract

Show Authors & Abstract
12:20 - 12:50
Predictive Crime Analytics Using Linear Regression and K means

Predictive analysis is concerned with the branch of data science used to predict future patterns and trends. This modelling technique can be used to aid society. In recent years, crime against women has skyrocketed and understanding past history can help to come up with insightful understandings that describe the current state of crime and assault in these countries.This research aims to foresee the crime patterns against women in India and USA. The studies are carried out in both countries to better understand the economic bearing, if any, in crimes committed against women. The data of the past years is studied using extensive EDA (Exploratory Data Analysis) techniques to help understand the problems womenface. The data is then normalized and Linear Regression is executed to predict future trends in crime rates. K means is then used to create clusters of states with the highest crime rates against women in India.This paper aims to help women and law enforcement by using technological advancements in the area of data science to predict future trends.
Authors: Chemmalar Selvi G (VIT Vellore), Preeti Rachel Jasper (VIT Vellore),
Hide Authors & Abstract

Show Authors & Abstract

Lunch Break 12:50 - 14:00

Technical Session 2 14:00 - 14:55

14:00 - 14:25
GLCM Feature Based Texture image classification using Support vector Machine

Texture is one of the important characteristics of an image and its features can be used to identify specific region of interest from the image. Texture features can be extracted for any kind of images like RGB, monochrome, aerial, satellite images. This paper describes an approach for texture image classification based on Gray Level cooccurrence matrix (GLCM) features and machine learning algorithm like Support vector machine. Ten different GLCM features were extracted and fed as input to Support vector machine for classification. The proposed method is trained and tested with dataset collected from center for image analysis, Swedish University. In recent days machine learning (ML) methods were highly used to mimic the complex mathematical expressions of texture features. The main objective of this paper is to demonstrate the state of the art of ML models in texture image prediction and to give insight into the most suitable models.
Authors: Anand R (Sona College of Technology, Salem, India), Shanthi T (Sona College of Technology, Salem), Paramasivam M E (Sona College of Technology), Sabeenian RS (Sona College of Technology), Manju K (Sona College of technology),
Hide Authors & Abstract

Show Authors & Abstract
14:25 - 14:55
Secured On Demand Adaptive Routing protocol for Data Transmission in IoT Environment

The emerging concept Internet of Things (IoT) has the capability to communicate data among devices throughout the entire world without human intervention. Existing reactive routing based protocol needs tremendous bandwidth, consumes more processing time and generates large number of control packets for communication. This paper proposes an efficient Secured On-demand Adaptive Routing (SOAR) protocol for IoT environment with secured data transmission among nodes. The proposed protocol is simulated and tested by using NS2 simulators. The simulation results show that throughput and packet delivery ratios are better than existing protocols.
Authors: P.Deepavathi - (National Institute of Technology, Trichy. Tamilnadu.), Dr.C.Mala - (National Institute of Technology, Tiruchirappalli, Tamilnadu.),
Hide Authors & Abstract

Show Authors & Abstract

Tea Break 14:55 - 15:00

Technical Session 3 15:00 - 16:50

15:00 - 15:25
Service Measurement Index based Cloud Service Selection using Order Preference by Similarity to Ideal Solution based on Intuitionistic Fuzzy Values

Abstract: Cloud Computing allows connection to a public resource pool on demand and easy network connection. Due to the popularity and profits of using Cloud Services, many organizations are moving to Cloud .So selecting a suitable and best Cloud Provider is a challenge for all the users. Many ranking approaches had been proposed for solving this multicriteria decision making problem like AHP, TOPSIS etc. But many of the works focused on quantitative QoS attributes .But qualitative attributes are also important in the case of many application scenarios where the user may be more concerned about the qualitative attributes. CSMIC has released Service Measurement Index attributes for effectively comparing the Cloud services. The comparison of Cloud Service providers based on SMI attributes which are qualitative in nature by using ranking approach is the objective of this paper. The proposed approach uses the MCDM algorithm called Technique for Order Preference by Similarity to ideal Solution and uncertainty is handled by Intuitionistic fuzzy values. The qualitative SMI attributes are used as criteria for ranking the Cloud Services.
Authors: Thasni T (Presidency University), C Kalaiarasan (Presidency University), K A Venkatesh (Myanmar Institute of Information Technology),
Hide Authors & Abstract

Show Authors & Abstract
15:25 - 15:55
A Multi-Objective Optimal Trajectory Planning for Autonomous Vehicles using Dragonfly Algorithm

Trajectory planning is considered a major challenge in autonomous driving, which faces significant issues like safety and efficiency. This paper proposes a novel swarm intelligence meta-heuristic optimization algorithm, called dragonfly algorithm, for the lane-change behavior in the navigation of autonomous vehicles. A multi-objective lane-changing trajectory planning method has been proposed to optimize the trajectory and avoid collision which mimics the dynamic and static swarm behaviors of the natural dragon-flies. Whenever the autonomous vehicle senses an obstacle, automatically the lane change maneuver should take place. The feasibility and the effectiveness of the algorithm are verified by simulation results using the lane-change data from the benchmark NGSIM dataset. Simulation results show that the proposed algorithm for trajectory planning gives an optimal path for lane change scenario considering both static and dynamic obstacles.
Authors: Syama R (National Institute of Technology, Tiruchirappalli, India), Mala C (National Institute of Technology, Tiruchirappalli, India),
Hide Authors & Abstract

Show Authors & Abstract
15:55 - 16:20
Emperor Penguin Optimization Algorithm and M-Tree-based multi-constraint Mul-ticast Ad hoc On-demand Distance Vector Routing Protocol for MANETs

Multicasting routing in ad hoc networks is considered essential for attaining reliable data dissemination. However, reliable data transmission can be achieved based on the estimation of optimal multicast trees that aid in better performance of the network. Further, prolonging network lifetime is yet another issue that needs to be concentrated for sustained connectivity. In this paper, Emperor Penguin Optimization Algorithm and M-Tree-based Multicast Ad hoc On-demand Distance Vector Routing (EPOA-MT-MAODV) Protocol is proposed for optimal selection of multicast routes for enhancing the lifetime of the network. This proposed EPOA-MT-MAODV protocol utilizes the merits of exploitation and exploration inherited from emperor penguin optimization algorithm with the estimation of multifactor, path inclusion and destination. It focuses on delay, minimum distance, link stability and energy for optimal selection of optimal tree. The simulation results of the proposed EPOA-MT-MAODV protocol confirm better performance in terms of energy consumption and Link Lifespan Time (LLT) for varying number of mobile nodes and mobility speed.
Authors: Deva Priya M (Department of Computer Science and Engineering, Sri Krishna College of Technology, Coimbatore, Tamilnadu, India), Rajkumar M. (Department of Information Technology, Sri Krishna College of Engineering and Technology, Coimbatore, Tamilnadu, India), Karthik S. (Department of Computer Science and Engineering, SRM Institute of Science and Technology, Delhi-NCR Campus, Ghaziabad, India), Christy Jeba Malar A. (Department of Information Technology, Sri Krishna College of Technology, Coimbatore, Tamilnadu, India), Kanmani R (Department of Information Technology, Sri Krishna College of Technology, Coimbatore, Tamilnadu, India), Sandhya G. (Department of Computer Science and Engineering, Sri Krishna College of Technology, Coimbatore, Tamilnadu, India), Anitha Rajakumari P. (Department of Computer Science and Engineering, SRM Institute of Science and Technology, Delhi-NCR Campus, Ghaziabad, India),
Hide Authors & Abstract

Show Authors & Abstract
16:20 - 16:50
Concurrent Spatial Color Information Processing for Video Based Vehicle Detection Applications

The increase in the number of vehicles requires an efficient and well-organized solution for traffic management and regulation. Among several traffic management system of Intelligent Transportation Systems (ITS), video based system has gained wide popularity that uses computer vision techniques. One of the significant process of computer vision technology is to detect vehicle from traffic videos. Effective image processing based computer vision techniques are available to detect vehicles. The accuracy of the detection primarily depend on the section of the video being processed for detection which is completely neglected by the existing image processing technique. Hence, in this paper a Concurrent Spatial Color Information Processing System (CSCIPS) is proposed that uses a novel detection framework known as Virtual Multi-Loop Crate (VMLC). The proposed framework VMLC uses all the spatial color information for image processing technique concurrently without loss of any information. The performance of the proposed CSCIPS system with VMLC is analysed with background subtraction based image processing techniques in benchmark traffic videos. The results show better performance in processing time, storage space and accuracy for vehicle detection when compared to other existing frameworks.
Authors: Manipriya Sankaranarayanan (National Institute of Technology Tiruchirappalli Tamil Nadu India), Mala C (National Institute of Technology Tiruchirappalli Tamil Nadu India), Samson Mathew (National Institute of Technology Tiruchirappalli Tamil Nadu India),
Hide Authors & Abstract

Show Authors & Abstract
Day 2 19/12/2020
Room #1

Keynote Speech 2 09:30 - 10:30

Dr. Manik Sharma

Tea Break 10:30 - 10:50

Technical Session 4 10:50 - 13:00

10:50 - 11:15
Food Demand Forecast For Online Food Delivery Service Using CatBoost Model

Online food ordering has been proven a great source for businesses from a wide range of sectors. By using an online food ordering system, you can get your food to be delivered to your door without consuming much time. For businesses operating in the food industry including restaurants, agriculture, and many others, accurate forecast is of crucial importance because of the unpredictable demand pattern. In several studies, the choice of an appropriate forecasting model remains a concerning point. In this context, this research aims to analyse the performance of the CatBoost Gradient boosting model for the prediction of raw materials for a meal delivery company that operates in multiple cities, having multiple centers.
Authors: Parvathi R (Vellore Institute of Technology,Chennai), Pattabiraman V (Vellore Institute of Technology,Chennai), Ansh pujara (Vellore Institute of Technology,Chennai),
Hide Authors & Abstract

Show Authors & Abstract
11:15 - 11:45
A Novel Coherent Architecture for Traffic Signal Management in Internet of Things

With recent developments in transportation, there is a rapid growth in vehicles leading to increase in traffic congestion. To manage this scenario, traffic management applications such as Traffic Signal Management (TSM), traffic congestion, route guidance, speed advisory, etc., based on Internet of Things (IoT) infrastructure is used. IoT is an overall system associating all the smart elements together and when confined with vehicles it is termes as Internet of Vehicles(IoV). The major challenge in IoV is the availability of information such as, traffic vehicle speed, density, volume headway, etc. for these applications during low propagation delay or high network utility. To resolve these challenges, this paper proposes a novel architecture namely Coherent Architecture for Traffic Signal Management in IoT (CATSMI) to manage traffic efficiently by improving and ensuring information availability. CATSMI uses recent advanced technology of Radio Frequency Identification (RFID), a non-intrusive sensors to support TSM developed using IoT. The performance of the proposed architecture is analysed using simulated values of IoV and RFID for TSM. The results show that the quality and precision of the traffic signal management application with IoV has improved using the proposed CATSMI.
Authors: Umaa Mageswari S (National Institute of Technology Tiruchirappalli, Tamil Nadu, India), Mala C (National Institute of Technology, Tiruchirappalli, Tamil Nadu, India), Santhanavijayan A (National Institute of Technology Tiruchirappalli, Tamil Nadu, India),
Hide Authors & Abstract

Show Authors & Abstract
11:45 - 12:20
Improved Rider Optimization Algorithm-based Link Aware Fault Detection (IROA-LAFD) scheme for securing Mobile Ad Hoc Networks (MANETs)

Securing communication in a dynamic network like Mobile Ad-hoc Network (MANET) is considered as the crucial task. The faulty links need to be detected for attaining better performance in terms of reliability and availability of mobile nodes in the network. Moreover, attacks need to be detected through a single stage attack detection process in a proactive way for guaranteeing quality of service in the network. However, the detection accuracy of most of the existing approaches of the literature still possesses a room for improvement. In this paper, Improved Rider Optimization Algorithm-based Link Aware Fault Detection (IROA-LAFD) scheme is proposed for facilitating security by resisting gray hole and black hole attacks with enhanced link stability. This IROA-LAFD scheme targets on mitigating packet dropping in a potential manner based on the steps that includes the discovery of neighbor and route, detection of attack, analysis of links and transmission of secure packets and link fault detection using IROA Algorithm. In particular, IROA is used for exploring the link paths established between the trust nodes of the network. The simulation experiments of the proposed IROA-LAFD scheme is identified to increase the packet delivery rate by 13.42%, throughput by 12.84% and the link fault detection rate by 15.48% when compared to the benchmarked schemes considered for investigation.
Authors: Sengathir Janakiraman (Department of Information Technology, CVR College of Engineering, Mangalpally, Vastunagar, Hyderabad), Deva Priya M (Department of Computer Science and Engineering, Sri Krishna College of Technology, Coimbatore, Tamilnadu, India), Aishwaryalakshmi G. (Lecturer, PSG Polytechnic College, Coimbatore, Tamilnadu, India), Suganya T. (Department of Computer Science and Engineering, Sri Krishna College of Technology, Coimbatore, Tamilnadu, India), Sam Peter S. (Department of Computer Science and Engineering, Sri Krishna College of Technology, Coimbatore, Tamilnadu, India), Karthik S. (Department of Computer Science and Engineering, SRM Institute of Science and Technology, Delhi-NCR Campus, Ghaziabad, India), Christy Jeba Malar A. (Department of Information Technology, Sri Krishna College of Technology, Coimbatore, Tamilnadu, India),
Hide Authors & Abstract

Show Authors & Abstract
12:20 - 12:45
Fog Assisted Real-Time Coronary Heart Disease Risk Detection in IoT Based Healthcare System

Advances in healthcare systems are helpful in the diagnosis and treatment of extremely critical diseases. Continuous monitoring of individuals leads to huge amount of medical data and new technology solutions are highly essential to handle the generated data. Fog computing approach is proposed in this paper to develop an efficient wearable device for wireless healthcare monitoring. Fog computing proves to be better than other remote -health monitoring methods because of its fast decision making capability and delivery of simple notifications to users. Centralized cloud server is utilized in fog computing framework for performing complex computations. In this work, machine learning based Coronary Heart Disease (CHD) risk detection is carried out with fog manipulation technique. Machine learning based CHD risk assessment is performed on HRV feature extracted ECG signal, blood pressure, glucose and cholesterol level. Wavelets are applied in this work for ECG signal feature extraction and identifying heart rate related parameters. Various machine learning algorithms such as decision tree classifier, SVM, ANFIS and KNN are utilized for classification. ANFIS classification is found to be effective in CHD risk prediction which categorized the given individual data into normal and CHD risky. Based on this work, it is possible to provide warning to the individuals to correct their food habits and lifestyle.
Authors: JUBAIR AHMED L (Sri Eshwar College of Engineering), ANISHFATHIMA B (Sri Krishna College of Engineering and Technology), GOKULAVASAN B (Sri Eshwar College of Engineering), MAHABOOB M (Sri Eshwar College of Engineering),
Hide Authors & Abstract

Show Authors & Abstract

Lunch Break 12:45 - 14:00

Technical Session 5 14:00 - 15:30

14:00 - 14:20
A Hybrid Algorithm for Document Clustering Using Optimized Kernel Matrix and Unsupervised Constraints

Document clustering is the most needed process in the data mining field where the number of documents with different methodologies are scattered. Meaningful information can be extracted from a collection of documents by grouping them effectively. There are various existing researches p which concentrate on clustering the documents. In the previous works, document clustering is done by using the methodologies like Term Weight based Hybridized Harmony K-Means (TW-HHKM) search and Coverage Factor based Hybridized Harmony K-Means (CF-HHKM) search. Clustering is normally done by using the K-means algorithm and the centroids of clusters are found optimally by using the harmony search algorithm. The main challenge faced by the existing methods is the reduced accuracy as unrelated documents may be grouped together. To overcome this problem, Novel Feature Weighting and Feature Selection based hybridization of optimized and unsupervised constraint kernel matrix k-means and Harmony Search method for Document Clustering (NFW-FS- HSDC) approach is introduced for accurate clustering of documents. The NFW-FS- HSDC algorithm optimizes the kernel matrix based on the feature ranges of dataset by utilizing PSO. The exploratory tests are conducted on News group and Trec dataset from which it is understood that the proposed NFW-FS- HSDC offers better accuracy values.
Authors: Siamala Devi S. (Department of Computer Science and Engineering, Sri Krishna College of Technology, Coimbatore, Tamilnadu, India), Deva Priya M (Department of Computer Science and Engineering, Sri Krishna College of Technology, Coimbatore, Tamilnadu, India), Anitha Rajakumari P. (Department of Computer Science and Engineering, SRM Institute of Science and Technology, Delhi-NCR Campus, Ghaziabad, India), Kanmani R. (Department of Information Technology, Sri Krishna College of Technology, Coimbatore, Tamilnadu, India), Poorani G. (Department of Computer Science and Engineering, Sri Krishna College of Technology, Coimbatore, Tamilnadu, India), Padmavathi S. (Department of Computer Science and Engineering, Sri Krishna College of Technology, Coimbatore, Tamilnadu, India), Niveditha G. (Department of Computer Science and Engineering, Sri Krishna College of Technology, Coimbatore, Tamilnadu, India),
Hide Authors & Abstract

Show Authors & Abstract
14:30 - 14:50
Media Access Protocol for Wireless Sensor Network using Active Reception Scheme Based Energy Efficient Technique

In this paper we propose Reception MAC (R-MAC), a time division multiple access (TDMA) based MAC protocol for data transmission in wireless sensor networks (WSNs). The power of R-MAC is to conserve energy through limiting idle listening, traffic overhearing, retransmission of data and packet collision. R-MAC uses a distributed scheduling procedure to assign a time slot for reception called Reception Slot (RS) to each sensor, and spreads information of assigned RS to each of its neighboring sensors. A sensor who have data to transmit can consequently wakes up in reception slot of its intended receiver only. R-MAC guarantees that only one one-hop neighbor can transmit at a time in reception slot of its intended receiver which assures collision avoidance. The performance of R-MAC is analysed in terms of offered traffic, power consumption and data delivery delay through detailed simulation. The performance of R-MAC is also analysed in terms of end-to-end delay and power consumption in multi-hop networks. The performance of R-MAC is compared against SMAC. Network simulator 2 (NS-2) is used for simulation.
Authors: Anushree Goud (NMIMS Mumbai), Dr. Bindu Garg (Bharati Vidyapeeth College of Engineering, Pune),
Hide Authors & Abstract

Show Authors & Abstract
15:00 - 15:20
Efficient Routing Strategies For Energy Management in Wireless Sensor Network

Wireless Sensor Network (WSN) refers to a group of distributed sensors that are used to examine and record the physical circumstances of the environment and coordinating the collected data at the centre of the location. This WSN plays a significant role in providing the needs of routing protocols. One of the important aspects of routing protocol in accordance with Wireless Sensor Network is that they should be efficient in the consumption of energy and have a prolonged life for the network. In modern times, routing protocol which is efficient in energy consumption is used for Wireless Sensor Network. The routing protocol which is efficient in energy consumption is categorized into four main steps: CM- Communication Model, Reliable Routing, Topology based routing, and NS- Network Structure. The network structure can be furtherly classified as Flat/hierarchical. The communication model can be furtherly classified as Query, Coherent/non-coherent, negotiation-based routing protocol system. The topology-based protocol can be furtherly classified as Mobile or location-based. Reliable routing can be fur-therly classified as QoS (Quality of service) or Multiple-path based. A survey on routing proto-col which is energy efficient on Wireless Sensor Network was also provided in this research.
Authors: Raviteja Kocherla (Research Scholar,CSE, Presidency University, Bangalore-560064, India), Ramesh Vatambeti (Associate Professor,CSE, Presidency University, Bangalore-560064, India),
Hide Authors & Abstract

Show Authors & Abstract

Closing Ceremony 15:30 - 16:00