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

Engineering and Technology

Global energy demands are growing rapidly. And, nonrenewable energy sources will not be able to meet our future energy requirement. Renewable energy, which has the benefit of minimum carbon emission, may be a feasible solution to make our planet safer and energy proficient. Over the past few years, many sorts of renewable energy resources such as wind, solar, geothermal, biomass, and tidal have been exploited. Computing and machine learning (ML) improves the potency and accessibility of renewable energy technology [1]. Computing (artificial intelligence [AI]), in conjunction with many AI advanced technologies, has incontestable immense potential to work on the renewable energy. AI and ML technologies will build a control by reducing emissions and increasing production potency. In this era, ML algorithms are applied to the information gathered from advanced sensors, good meters, intelligent device, and grid operators to estimate how individual appliances behave. Germany, as an example, uses advanced AI technology in the early warning system, which takes time period information from wind turbines and star panels around the country to predict the energy report for next two days. AI may also facilitate the trade to improve safety, responsibleness, and potency. It may also offer visibility into energy run, consumption patterns, and instrumentation health. As an example, prophetic analytics will take device knowledge from a turbine to watch wear and tear, and predict with a high degree of accuracy once it might want maintenance.

School of Computer Engineering & Technology

Artificial Intelligence and Machine Learning Methods for Renewable Energy

Misalignment is probably the main cause for failure of the rotating machinery. Due to misalignment, excessive stresses and vibrations arise in the system which cause failure of different components of rotating system. The present work is therefore taken up to predict the unique vibration spectrum characteristics for diagnosing parallel misalignment at varying operating conditions and for different geometrical parameters. In the present study, the work has been carried out to establish relationship between parallel misalignment and harmonic frequency components of its vibration signature by experimental and numerical method. An innovative experimental set-up with provisions for changing geometrical as well as operational parameters has been developed for the study of parallel misalignment. To set precise predetermined parallel misalignment values the compound slide table with rotating calibrated wheels was used along with the latest laser alignment system. It was noticed that, in the prediction of misalignment, the knowledge of natural frequency of the system is essential. The natural frequency of the setup is determined experimentally using modal analysis. The significant characteristic frequency components obtained from vibration signature are correlated with resonance frequency.

School of Mechanical Engineering

SAE Technical Paper

Experimental investigation of parallel misalignment using vibration signature analysis

With an increase in defined carrier frequency for the 5G new radio system, the need for synchronization also increases. The transceiver loses synchronization due to the occurrence of timing and carrier frequency offsets. Carrier frequency offsets often occur due to mismatch between transmitter and receiver oscillator frequency as well as the occurrence of doppler shifts due to transmitter/receiver movements. When frequency offsets exceed subcarrier spacing, integer frequency offset occurs that results in performance loss due to subcarrier indices shifts. Conventional approach i.e. maximum likelihood and sequential method is already employed to estimate integer frequency offset and to detect sector id. In this paper, the deep learning-based method is demonstrated to estimate the integer frequency offset and sector id detection. The neural network containing multiple convolution layers with activation layers is used to find the optimum received signal. Then, by calculating the number of cyclic shifts in the optimum received signal, the integer frequency offset is estimated. Using the corrected optimum received signal, the primary synchronization signal is also detected that gives sector id. This proposed estimator is tested for different profiles of tapped delay line models with different desired delay spread and compared with conventional methods i.e. maximum likelihood estimation method and sequential estimation method. Simulation results show that the proposed Neural Network based estimator outperforms in all delay profiles.

School of Electronics & Communication Engineering

International Journal of Intelligent Engineering and Systems

A Neural Network based Integer Frequency Offset Estimation and PSS Detection in 5G NR Systems

Annotation of images is a time-consuming task. In recent times captioning of images is required in many Artificial Intelligence applications. Captioning Images intends to produce a sound and thorough portrayal that sums up the contents of a picture. Traditional approaches cannot handle the complexity and difficulties of image captioning as well as deep learning-based approaches. In this paper an encoder-decoder version with an attention mechanism integrating the picture feature that generates a semantic and syntactically sound description. One of the available benchmark datasets, Flickr8k is used to evaluate the model. The system incorporates an EfficientNet B3 encoder that draws out features from an image. It uses the mapped caption of images to produce the corpus. For the caption generation process, the model implements the attention and Gated Recurrent Unit (GRU) model to effectively bring about the image description. The various cosine similarity and N-Gram feature extraction means have been used for generating the blue score for the entire testing dataset and show the effectiveness of the proposed system.

Image Captioning using Deep Learning based System

Image Captioning using Deep Learning based System

Deep learning in computer vision has shown remarkable success in the performance of detection systems for plant diseases. However, due to the complexity and deeply nested structure of these models, these are still considered as black-box and explanations are not intuitive for human users. Many researchers have developed deep neural architectures for plant disease detection but have not provided classification explanations. To be used in practical applications, our model needs to explain why the model classified a given image. Explainable Artificial Intelligence (XAI) provides algorithms that can generate human-understandable explanations of AI decisions. In this paper, we summarize recent developments in XAI techniques, develop a plant disease detection system, and most importantly an explainable AI method named Gradient-weighted Class Activation Mapping ++ (GradCAM++) is used to locate the disease and highlight the most important regions on the leaves contributing towards the classification.

Explainable AI for Deep Learning Based Disease Detection

School of Civil Engineering

Lecture Notes in Civil Engineering

An experimental study on PET bottle mattress as reinforcement on fly ash fills

Reading is a fundamental part of our lives. Information and text is present everywhere in magazines, in newspapers, in receipts, in letters, bank documents etc. But text present on medicine and its cover is something which is important and needs to be crosschecked every time we buy it. While buying medicines, people tend to make sure that they are buying a correct one by cross checking the medicine name, expiry date, manufacturing date and the MRP so that they do not get fooled. It is one of the easiest tasks to perform for normal people as they can read. But what about the people with visual impairment? How will they be able to get their medicine details? The proposed application tries to imitate this job for them and helps them better understand their medicines. The application scans the medicine content side and uses the advanced technologies like Tesseract optical character recognition (OCR), Computer Vision, Text to Speech technology to extract the required information and makes it available to the user in the form of audio. The application helps visually impaired people to independently carry out this task without the hindrance of a third person.

Turkish Journal of Computer and Mathematics Education (TURCOMAT)

<a href="https://www.turcomat.org/index.php/turkbilmat/article/download/10346/7811">Medicine Identification Application for Visually Impaired People</a>

The brain tumor is among the most hazardous and destructive diseases. The mortality rate in brain cancer is more at a later stage. Also, the brain tumor's misdiagnosis will produce danger and reduce the patient's chances of survival. The early diagnosis of a brain tumor aids in saving the life of the affected person by providing the right treatment. The computer-aided medical imaging techniques like Magnetic Resonance Imaging (MRI) and Computed Tomography (CT) aids in diagnosing the disease. Hence brain MRI classification became an active research area in recent years. Numerous methods have been presented in the earlier period for MRI categorization, right from classical methods to advance Deep Learning (DL) algorithms such as Convolutional Neural Network (CNN). The conventional Machine Learning (ML) techniques required handcrafted features while CNN performs classification by drawing features from unprocessed images directly via the convolution and pooling layer's parameter tuning. The feature extraction using the CNN algorithm is mostly influenced by the size of the training process's images. CNN models overfit after some epoch if the training dataset size is small. Therefore, transfer learning techniques have evolved. In the proposed system, conduct five studies using five transfer learning architectures such as AlexNet, Vgg16, ResNet18, ResNet50, and GoogLeNet to classify the clinical dataset of brain MRI into benign and malignant. Data augmentation techniques are applied over the brain MRI to generalize the results and reduce overfitting possibilities. In this proposed system, fine-tuned AlexNet architecture achieved the highest precision, recall, and f-measure value of 0.937, 1, and 0.96774, respectively

Journal of Engineering Science and Technology (JESTEC)

COMPARATIVE ANALYSIS OF PERFORMANCE OF DEEP CNN BASED FRAMEWORK FOR BRAIN MRI CLASSIFICATION USING TRANSFER LEARNING

Fulltext

Indian Journal of Science and Technology

Strength and Swelling Behavior of Black Cotton Soil Stabilized with Waste Rubber Tyre

Smart Campus is approach that combines the modern technology with the human innovation in order to use resources efficiently. There are many different individual smart solutions which altogether makes smart campus with some approaches, in which the main approach is the smart parking. Locating parking space in campus, especially during the peak hours, is clumsy for drivers. We have developed a prototype to ease the problem of parking in the campus. This project comprises of an on-site placement of Internet of Things (IOT) module that is used to look after and provide the status of availability of every single parking spot which helps the user to locate the closest parking spot and gives information of available parking spots in that respective parking lot. This system uses functional sensors in parking area which are used to get the real time status of the parking using the deployed active data processing units. To get the real time available slots we have used the Google firebase real-time database. The Django web framework is used to create the web application so that user can get the information of the free available parking slot in the campus using Wi-Fi or Internet. Proposed approach mostly focuses on controlling the time in searching the parking spots and it also avoids the additional travelling through full parking spots in a parking area.