http://nopr.niscpr.res.in/handle/123456789/57968 2026-10-09T19:43:22Z http://nopr.niscpr.res.in/handle/123456789/57978 Title: VISU: A 3D Printed Functional Robot for Human Pose Replication Authors: Kompally, Pranav; Sethuraman, Sibi Chakkaravarthy; Reddy, Srikar; Koduri, Charan; Naidu, Yaswanth; Raghavaiah, M; Reddy, Sashidhar; Nikhil, Namburi Abstract: This paper presents VISU, a novel 3D printed functional robot. VISU is equipped with open-source technologies making it more modular in adapting Internet of Things (IoT) based services. VISU is able to detect and analyze the user’s activity and pose. In addition, a simple method to replicate the pose of a user is also proposed. VISU can also perform actions such as Face recognition, Object Recognition among other basic functionalities. Page(s): 563-569 2021-07-01T00:00:00Z http://nopr.niscpr.res.in/handle/123456789/57977 Title: Knitting Machinery Spare Classification using Deep Learning with Differential Privacy Authors: Tastimur, Canan; Kasap, Songul; Akin, Erhan Abstract: Given their widespread use, knitting machines must be maintained regularly. When the spare parts that make up these machines break down or become unusable, they must be replaced with new ones. However, the code/name information of the spare parts is not available to the end user, and can only be accessed with high-cost catalog procurement. Manufacturing companies keep the code/name information of such machine parts confidential. When the literature is examined, there are no studies in which spare parts are classified with machine learning–based algorithms. In line with this, this study focuses on the classification of spare parts using machine learning–based algorithms. The deep learning–based Convolutional Neural Network (CNN) architecture developed in this study can classify highly similar spare parts. In addition, since the code/name information received from the manufacturer and the spare part sample images require confidentiality, the CNN architecture has been developed in combination with the Differential Privacy (DP) method to present the DP-CNN method. As a result of the application of the Differential Privacy method, there has been no great loss of accuracy. This is an important development for our study. In the article, many optimizer algorithms are tested on the proposed method and comparative results are given. A 99.41% accuracy ratio has been obtained with the DP-RMSProp optimization method, which produces the best results. Experimental results of our study are presented in detail. Page(s): 570-581 2021-07-01T00:00:00Z http://nopr.niscpr.res.in/handle/123456789/57976 Title: Multi-Objective ANT Lion Optimization Algorithm Based Mutant Test Case Selection for Regression Testing Authors: Tripathi, Aprna; Srivastava, Shilpa; Mittal, Himani; Sinha, Shivaji; Yadav, Vikash Abstract: The regression testing is principally carried out on modified parts of the programs. The quality of programs is the only concern of regression testing in the case of produced software. Main challenges to select mutant test cases are related to the affected classes. In software regression testing, the identification of optimal mutant test case is another challenge. In this research work, an evolutionary approach multi objective ant-lion optimization (MOALO) is proposed to identify optimal mutant test cases. The selection of mutant test cases is processed as multi objective enhancement problem and these will solve through MOALO algorithm. Optimal identification of mutant test cases is carried out by using the above algorithm which also enhances the regression testing efficiency. The proposed MOALO methods are implemented and tested using the Mat Lab software platform. On considering the populace size of 100, at that point the fitness estimation of the proposed framework, NSGA, MPSO, and GA are 3, 2.4, 1, and 0.3 respectively. The benefits and efficiencies of proposed methods are compared with random testing and existing works utilizing NSGA-II, MPSO, genetic algorithms in considerations of test effort, mutation score, fitness value, and time of execution. It is found that the execution times of MOALO, NSGA, MPSO, and GA are 2.8, 5, 6.5, and 7.8 respectively. Finally, it is observed that MOALO has higher fitness estimation with least execution time which indicates that MOALO methods provide better results in regression testing. Page(s): 582-592 2021-07-01T00:00:00Z http://nopr.niscpr.res.in/handle/123456789/57975 Title: Classification of Addiction Behavior based on Regular and Rare Model Authors: Sabapathi, V; Peter, J Selvin Paul Abstract: Realization is the comprehension of existence in its widest terms. Many of us, both physically and virtually, are unconscious of our level of addictive concern. Predicting virtual and emotional-based activity poses certain difficulties in determining an addiction level. Specifically, how to compute the addictive and what types of controls can help us monitor the addiction and get a good estimate of the individual's addicted stage. The threshold levels vary depending on a variety of factors such as age, gender, society, and so on. The addiction mentality system's prediction plays a vital role. In this regard, our research develops a Regular and Rare (RAR) based classification model for finding effective addiction predictors. This RAR classification and prediction technique is based on an examination of addiction patterns' consistency. This strategy focuses on the length of time spent doing the same activity rather than the amount of quantity consumed. The concept behind it if an individual consumes a low density of nicotine but persists for a decade, this is considered as a habitual and addictive activity. In such a way that if an individual doesn't really engage in the very same type of activity for an extended period of time, the action may be considered an uncommon occurrences rather than an addictive class. Page(s): 593-599 2021-07-01T00:00:00Z