MIT-World Peace University (MIT-WPU) is a top educational institute located in Pune, India. It is recognized within&nbsp;UGC under Government of Maharashtra Act No. XXXV 2017.&nbsp;

It is part of the MIT group of institutions (established in 1983) which covers 10+ campuses across India and has more than 50,000+ students enrolled across various disciplines at a given point of time. The institute houses 15 schools ranging from Engineering &amp; Technology, Design, Pharmacy, Commerce, Governance among others. It offers undergraduate, postgraduate and PhD programs among different disciplines.&nbsp;

&nbsp;

MIT World Peace University

School of Computer Engineering & Technology

The hand sign language uses visual-manual modality to share a certain message. Specially abled people having hearing and speaking disabilities interact more naturally in hand sign language rather than verbal language. According to one of the Census study, 2.21% out of 121 crore population in India are 'disabled', out of which 19% are having a hearing disability and 7% are having speech disability. Since everyone cannot communicate in sign language as it is a lesser-known language, it often leads to communication gap. So, the automated Sign Language Interpreter (SLI) helps to meet this communication gap as a manual sign language translator is not a convenient option because of its privacy issues and lack of availability. This paper proposes an Indian Hand Sign Language Interpreter which operates upon a vision-based approach that uses Machine Learning and Deep Learning techniques to locate the hand gesture region accurately for extracting the features and finally interpreting the respective meaning. The experimentation for performance metrics such as accuracy and loss using various activation functions helped to analyzed the performance of the model. The system successfully identifies a number of hand spelled words and thus eases the communication among people. © 2021 IEEE.

World University root node

Engineering and Technology

2021 IEEE Pune Section International Conference, PuneCon 2021

Interpretation of Hand Spelled Banking Helpdesk Terms for Deaf and Dumb Using Deep Learning

At present, India has the largest population of below 14 years children in the Asia Pacific. With the increasing birth rate, critical Pneumonia cases have been referred to Neonatal Hospital for treatment. As the number of adults in India who have tested positive for COVID-19 has grown, so has the number of children who have contracted it. However, we haven't noticed a dramatic increase in the number of children infected with COVID-19 across the country. It's important to note that, unlike the previous wave, the second wave is more likely to infect whole homes. We must be vigilant and adhere to COVID-19's recommended practices. The current study says that the mortality rate of Pneumonia and Covid-19 infection in rural areas is high. Radiology plays a vital role to diagnose Pneumonia by the examination of X-Ray images. Over the years CNN Architectures have evolved and now produce appreciable accuracy (over 85%) for classification tasks. This has promoted the use of CNN Architectures in the field of medicine, especially for classification tasks such as disease detection from x-rays. This implementation evaluates the performance of four popular CNN Architectures viz. VGG16, ResNet50V2, InceptionV3 and MobileNetV2. The implementation will classify x-ray images into normal, covid, and pneumonia and then compare the performance of the aforementioned models over the accuracy, Area under the curve (AUC), precision, recall metrics. © 2021 IEEE.

Proceedings of the IEEE International Conference Image Information Processing

Comparison and Evaluation of CNN Architectures for Classification of Covid-19 and Pneumonia

Human beings when born are adorned with a presence that helps them to differentiate among other individuals around ie. 'Face'. Alongside speech and fingerprints, A face is significantly used in biometrics as no two individuals have the same face. Myths though say that in this world you may resemble with some facial characteristics to 7 different individuals then these facial detection and recognition algorithms help us to identify between them. Facial expressions detection, face detection, and recognition have been impactful in the research domain especially in fields of image processing. Although it might sound like a cakewalk though undoubtedly it has proven to be a complex task for a computer to detect and pose all the features of the image passed since various other variables impair the accuracy/confidence of the algorithms mainly illumination variation, the background noise of the image, low resolution, occlusion, blurriness of the image, pixel scale of the image and many more components.Considering these myths, this research is based upon face recognition algorithms and a model to identify if there is a face in an image passed, identify the number of faces, and justify with confidence value obtained which algorithm should be suggested, is most optimum for face detection and recognition among (Haar classifiers, default face recognizer, Local Binary Pattern Histogram (LBPH) face recognizers, {Adaboost algorithm, Eigenface methods, etc.). This research particularly focuses on the use of Haar cascades and the LBPH Face recognizer algorithm as a hybrid combination for the smart voting system as an authentication approach. © 2021 IEEE.}, author_keywords={Face detection; Face Recognition; Haar cascade Classifier; Image processing; LBPH Face Recognizer; Smart Voting system}, keywords={Adaptive boosting; Image classification; Voting machines, Face detection and recognition; Face recognition systems; Faces detection; Haar cascade classifiers; Haar classifiers; Images processing; Local binary pattern histogram face recognizer; Local binary patterns; Smart voting system; Voting systems, Face recognition}, publisher={Institute of Electrical and Electronics Engineers Inc.}, isbn={9781665425032}, language={English}, abbrev_source_title={Int. Conf. Smart Gener. Comput., Commun. Netw., SMART GENCON}, document_type={Conference Paper}, source={Scopus},

2021 International Conference on Smart Generation Computing, Communication and Networking, SMART GENCON 2021

Optimal Face Recognition System using Haar Classifier

The most common psychiatric disorder is Clinical Depression. More than fifteen percent of people undergo an incident of major depression sometime during their life. There is an increase in depressive disorders and other-regarding symptoms in past years, which leads to the importance of early detection of depression. Detection of depression can be automated by analyzing a person's behaviour and emotions. The depressive disorder influences auditory speech qualities, facial expressions, thinking, and some cardiovascular activities; hence depression can be recognized by analyzing such factors and also trying to find out which method is more suitable and reliable. Through preprocessing, features of different signals are extracted and further handled to develop a machine learning model. Neural Networks are a faster and efficient way to analyze depression. This survey briefs about different techniques and machine learning models for depression detection. © 2021 IEEE.

2021 12th International Conference on Computing Communication and Networking Technologies, ICCCNT 2021

A Survey: Feature Extraction Techniques and machine learning models for Depression Analysis

With the recent growth of media platforms such as YouTube, Spotify, etc. These Platforms make content easily accessible but at the same time causes many problems like plagiarized content. Especially with audio content we can observe multiple types of plagiarism like unauthorized use of music for making remix music. To deal with this problem we use audio fingerprinting. It is the process of presenting an audio signal in a closed-packed way by extracting relevant characteristics of the audio signal. This in itself can be used to check the song, but it won't give an accurate result if the plagiarized song is copied from multiple different songs. To solve this particular problem, we use segment matching. © 2021 IEEE.

Proceedings - 1st International Conference on Smart Technologies Communication and Robotics, STCR 2021

Music Plagiarism Detection using Audio Fingerprinting and Segment Matching

Garbage disposal is a very important task in a healthy and green environment. With spreading awareness among the citizens of India regarding the importance of a clean environment to decrease the consumption of natural resources and garbage disposal, the recycling industry is booming. The quantity of generated trash in day-to-day life is affecting land, water, and air which causes a serious threat to the aquatic species and their surroundings and ultimately to humans if not managed properly. Conventional garbage disposal systems are on the rise and need accurate and efficient segmentation and recognition mechanisms. This demand coincides with the increase of computational capabilities of modern computer architectures and more effective algorithms for image recognition. To deal with this problem the garbage classification process is automated by building an image classifier using a convolutional neural network (CNN) and thereby decrease the time for the waste segregation and make it cost-effective. The idea behind making the process automated is to decrease human intervention and make this waste segregation process more productive. In this work, three different models are being tested for higher accuracy: Simple CNN, ResNet50, and VGG16 trained on various datasets of images, these are used to extract features from images and feed them into a classifier for dump/trash classification. The trained models are then fed to a mobile application that captures pictures in the real-time using camera. The experimental results conclude the performance of VGG16 using Transfer Learning being significantly higher than all other models for the purpose of Trash Classification and the performance of Simple CNN being better for Dump Classification. © 2021 IEEE.

2021 6th International Conference on Recent Trends on Electronics, Information, Communication and Technology, RTEICT 2021

Garbage Classifying Application Using Deep Learning Techniques

Enterprise optimization has become a major target in the pharmaceutical industry because of the increasing pressure for staying relevant in the worldwide commercial center. The ever-increasing competition, changing regulatory requirements, new emerging technologies along with a lot of other factors are pushing pharmaceutical manufacturers towards Enterprise optimization for the purpose of fast-Tracking innovation to boost their productivity and product quality. The objectives of Enterprise optimization include profit maximization, asset utilization, management of inventory levels, and coordinated decision making. These business exercises are moderately troublesome because of the association of various risk factors. In this paper, applications that have relevance to Enterprise optimization have been reviewed, with special attention to innovative drug development life cycles which are prone to substantial risks. This paper proposes a decision-making methodology with the identification of uncertainties and risks based on the Fuzzy AHP approach. © 2021 IEEE.

Enterprise Optimization in Pharmaceutical Industry

Advances in the domain of software-based technology pave the way for widespread use of object-oriented programs. There is a need to develop a well-established software system that will reduce maintenance costs and enhance the usability of the component. While designing a software system, its internal structure deteriorates due to prolonged or delayed maintenance activities. In such situations, restructuring of the software is a superior approach to improve the structure without changing external behaviour of the system. One approach to carry out restructuring is to use refactoring on the existing source code for improving the internal structure of the code. Code refactoring is an effective technique for software development that improves the software’s internal structure without changing its external behaviour. The purpose of refactoring is to improve the cohesion of existing code and minimize coupling in the existing module of a software system. Among numerous methods, clustering is one of the effective approaches to increase the cohesion of the system. Hence in this paper, the authors suggest to extract member functions and member variables and propose to find their similarity by Frequent Usage Pattern approach. Next, the proposed fuzzy based clustering algorithm perform effective code refactoring. The proposed method utilizes multiple refactoring methods to increase the cohesion of the component without any change in the meaning of the software system. The proposed system will offer automated support to change low-cohesive to high-cohesive functions. Finally, the proposed model has been experimentally tested with object-oriented programs. © Bobde S., Phalnikar R., 2021.

Fulltext

Scientific and Technical Journal of Information Technologies, Mechanics and Optics

Software restructuring models for object oriented programming languages using the fuzzy based clustering algorithm

Brahmi script is an ancient Indian script that requires expertise to read the written documents. OCR system can help reading and understanding the documents of Brahmi script. In the OCR systems, to recognize the symbols, accurate segmentation in earlier phase is required. If the segmentation goes wrong, it indicates the failure of the recognition process. Identification and segmentation of touching characters is a challenging task in OCR systems. In this paper, the authors have proposed an algorithm which successfully identifies and segments the touching characters from the Brahmi script documents. The algorithm uses ensemble classification technique to identify the touching characters using 9 unique features. The proposed algorithm has achieved 100% accuracy in identification and 99.16% accuracy in segmentation of touching Brahmi script characters. The results of proposed algorithm are significant to identify and correctly segment the touching Brahmi script characters from degraded digital estampage images. © 2021, Engg Journals Publications. All rights reserved.

Indian Journal of Computer Science and Engineering

IDENTIFICATION AND SEGMENTATION OF TOUCHING BRAHMI CHARACTERS FROM DEGRADED DIGITAL ESTAMPAGE IMAGES USING ENSEMBLE CLASSIFIER

Depression is a common and severe medical condition which adversely affects your feeling, thinking and acting. Depression can lead to suicide. It can trigger a slew of physical and emotional issues, as well as a reduction in your ability to function at work and home. Globally, depression affects more than 264 million individuals of all ages [1]. Since the younger generation is more reliant on social media, analyzing posts on social media will benefit in detecting depression. This research work has proposed a system to detect depression using ensembled learning and Natural Language Processing (NLP) techniques. Also, the proposed research work has compared the performance of multiple machine learning algorithms and the best performing configuration gave us the accuracy of 96.35%.