Forest fires in the Nilgiris, a fragile part of the Western Ghats, continue to damage forests, wildlife, and nearby settlements. Each fire season brings a new wave of challenges as changing weather and human activity continue to increase the risk. Past studies have produced fire-risk or vulnerability maps with the help of Remote Sensing and GIS, but these are more like snapshots; they do not tell us how the risk changes from day to day. To address this limitation, the present work focuses on building a system that can predict forest fire risk every day using multiple layers of data. The approach combines vegetation indices from satellites, elevation and slope data, land-cover information, daily weather factors like temperature and rainfall, and records of past fire events to form a detailed understanding of fire-prone zones. Proposed framework combines different sources of information as satellite vegetation indices, terrain and land-cover, daily weather readings, proximity to human activity and past fire records and arranges them in a spatio-temporal grid. Standard machine-learning models such as Random Forest and XGBoost are used as baselines, while image features and temporal patterns are explored with deep networks like CNNs and ConvLSTMs. To make the results transparent, explainability methods are applied which reveal how stress in vegetation, dryness, high temperature, slope, or settlement closeness contribute to risk. The outcome is not just another map but an early-warning tool that produces probability layers and risk categories through an interactive dashboard, designed for managers in the Nilgiris and adaptable to other Western Ghats regions.

Introduction Mucormycosis (black fungal attack) has recently been identified as a significant threat, specifically to patients who have recovered from coronavirus infection. This fungus enters the body through the nose and first infects the lungs but can affect other body parts, such as the eye and brain, resulting in vision loss and death. Early detection through lung CT scans is crucial for reliable treatment planning and management. Methods To combat the above problems, this paper introduces a Condition Generative Adversarial Network Deep Learning Model (CGAN-DLM) to facilitate the automatic lung CT image segmentation process, contributing to accurately identifying Mucormycosis earlier. This deep learning model employed different pre-processing strategies over raw lung CT images for extracting its ground truth values based on potential morphological operations. It adopted CGAN to segment the region of interest used for diagnosing mucormycosis with the pre-processed images and their related truth values. Results It also included a volumetric assessment approach that significantly identified the change in lung nodule size before and after the infection of mucormycosis. Conclusion The extensive experiments of the proposed CGAN-DLM conducted using lung CT images taken from the LIDC-IDRI database confirmed sensitivity of 98.42%, specificity of 98.86% and dice coefficient index of 97.31%, on par with the benchmarked lung CT images-based Mucormycosis detection approaches.

Aquaponics is a sustainable farming method that combines aquaculture and hydroponics to grow plants and fish in a closed-loop system. In this research paper, an irrigation system based on aquaponics is proposed, which uses real-time sensor data from the fish tank and crop soil to improve the efficiency of the system. The system is designed to make informed decisions about crop irrigation needs by visualizing the data for analytics. The study compares the accuracy of three classification algorithms, KNN, Naive Bayes, and ANN, to decide when to irrigate the soil based on real-time sensor data. The proposed irrigation system includes two sets of sensors, one for the fish tank and the other for the crop soil, which is processed by an Arduino board and sent to Adafruit’s cloud platform for visualization and analytics. This cloud-based platform allows easy access to real-time data, enabling efficient monitoring and control of the irrigation system. Additionally, the study visualizes the results obtained from using regular water and lake water in the aquaponics system.

Purpose the problem of big data analytics and health care support systems are analyzed. There exist several techniques in supporting such analytics and robust support systems; still, they suffer to achieve higher performance in disease prediction and generating the analysis.. For a hospital unit, maintaining such massive data becomes a headache. However, still, the big data can be accessed towards analyzing the bio signals obtained from the human body for the detection and prediction of various diseases. To overcome the deficiency, an efficient Health Care Big Data Analytics Model (HCBDA) is presented, which maintains a massive volume of data in the data server. Methods The HCBDA model monitors the patients for their current state in their cardiac and anatomic conditions to predict the diseases and risks. To perform analysis on health care, the model has accessed the data location by discovering the possible routes to reach the source. The monitored results on blood pressure, temperature, and blood sugar are transferred through the list of routes available. The network is constructed with a list of sensor nodes and Internet of Things (IoT) devices, where the sensor attached to the patient initiates the transmission with the monitored results. The monitored results are transferred through the number of intermediate nodes to the monitoring system, which accesses the big data to generate intelligence. The route selection is performed according to the value of Trusted Forwarding Weight (TFW) and Trusted Carrier Weight (TCW). At each reception, the features from the packet are extracted, and obtained values are fed to the decisive support system. The decisive support system cluster the big data using the FDS clustering algorithm, and the classification is performed by measuring the feature disease class similarity (FDCS). According to the class identified, the method would calculate Disease Prone Weight (DPW) to generate recommendations to the medical practitioner. Results The unique Health Care Big Data Analytics (HCBDA) paradigm for patient-centered healthcare using wireless sensor networks and IoT devices was described. The patient's bio signals are watched in order to provide medical assistance. In comparison to the previous methods, the proposed approach helps to generate higher performance in disease prediction accuracy up to 96%. Conclusion The value of Trusted Forwarding Weight (TFW) and Trusted Carrier Weight is used to determine the route (TCW). Sensor based IoT values like Pressure glucose, pulse oximeter, and temperature etc. the following parameters like classification accuracy and false ratio are calculated based on efficient machine learning model. The crucial support system receives the values it receives after each reception together with the features that were derived from the packet. The classification is carried out by calculating the Feature Disease Class Similarity, and the decision support system clusters the huge data using the FDS clustering technique.

Cardiac disease analysis in big data is an emerging factor for human health protection against heart attacks. Most cardiovascular diseases lead to heart failure due to an imbalance of immunity and attention in health conditions. Hence, immunity-based feature analysis of patients’ records is essential to predict accurate results. The machine learning methods make predictions depending on the extended-lasting features to analyze the health data. But the marginal features expose non-relational feature observation to reduce the classification prediction accuracy. We propose a Deep Spectral Time-Variant Feature Analytic Model (DSTV-FAM) using SoftMax Recurrent Neural Network (SMRNN) in a wireless sensor network to improve cardiac disease prediction accuracy. Initially, the IoT sensor devices collect the data from patient observation to validate the data transmission in route propagation. The collected data is organized as features in the collective dataset. The parts are initially preprocessed into the redundant dataset and estimate the Cardiac Immunity Influence Rate (CIIR) depending on the time-variant feature selection model. The estimated weights are marginalized as spectral features trained into the classifiers. Further, Soft-Max Activation Function (SMAF) creates a logical function depending on the Cardiac Affection Rate (CAR). Then the trained, rational neurons are constructed into a Recurrent Neural Network (RNN) Feed-forward feature values using a classifier and Rate of Disease Affection (RDA) by Class Type. The proposed structure yields high prescient exactness concerning order, accuracy, and review to help early treatment for early cardiovascular gamble expectation.

A brain tumor is an anomalous blowup of cells if not dually detected. A brain excrescence must be found as soon as possible for therapeutic use and survival chances. There are numerous different types of brain tumors with different features, sizes, and treatment choices. Tumor detection manually is difficult, laborious, and prone to mistakes. In this research, deep features are extracted and passed to an ensemble system of AlexNet, EfficientNet, ResNet50, InceptionV3, and VGG16 from which a score vector is acquired from Softmax for demarcation between glioma, meningioma, pituitary, and no tumor. The issues in the published studies have already been estimated on a standard dataset such as Kaggle, 2020-BRATS. The model has detection scores of more than 96%, which shows that the proposed ensemble model outperforms existing works.

People get to know one another by sharing their ideas, thoughts, and experiences with those in their immediate surroundings. There are numerous methods for accomplishing this, the most effective of which is the gift of “Speech.” Speech enables all people to communicate their ideas effectively and to comprehend one another. It will be unfair if we fail to take into consideration those who are denied this priceless gift: the dumb and deaf. The preferred method of communication in these situations has continued to be human hand contact. Things that have been first challenging or unattainable for people with disabilities are now regularly available to them and can be accessed by them with ease. Artificial intelligence made it possible for people with disabilities to live in a society where their challenges are acknowledged and taken into account (AI). Technological advancements have made it possible for technology to adjust and transform the world into a more open community. There is a certain sense of being human as AI directly correlates individuals, including people with and without impairments. hearing voice in the preferred language so that a message can be delivered it to normal people to build a model that is trained on various hand motions we are using a convolution neural network and deep learning on the basis of this model an app is created with the help of this app person who is deaf or dumb can communicate using postures that are translated into speech and human-understandable words

License plate recognition are used in toll plaza, surveillance cameras, intelligent car parking, etc,. This paper proposes three modules for number plate recognition: Image acquisition, License plate detection and Character recognition. Firstly, a pytorch library OpenCV is used for retrieving the data. YOLOv5, a family of You Only Look Once (YOLO) model is used for detecting the number plate. Finally, OCR methods i.e., Tesseract OCR and EasyOCR are used for recognizing and extracting the characters from the number plates. A dataset from github is used for training and testing the proposed model. The result shows that EasyOCR has resulted in more than 95% accuracy for predicting the number plate when compared to Tesseract OCR which has only resulted in 90% accuracy. Hence, EasyOCR outperforms Tesseract OCR as it uses deep learning approach for object recognition and it is efficient in real time prediction.

There is a great need to create and put in place a method of automatic detection as a substitute for conventional diagnosis for COVID-19 detection that can be employed on a commercialscale because there aren't as many COVID-19 test kits availablein medical institutions. In particular, chest X-Ray scans can beexamined to assess whether a patient has COVID. Due to the availability of numerous big annotated picture datasets, convolutional neural networks have achieved remarkable success in image analysis and classification. Input is obtained in the form of chest x-rays images. Output results are acquired instantly in real-time which predicts if the person suffers from Covid or not. Modern technique use the RCNN algorithm, which makes them less precise and time-consuming. We suggest an automated deep learning-base method for extracting COVID-19 from chest X-ray pictures. For analysing the chest X-Ray pictures, suggested method offers enhanced depth-wise convolution neural network. Through wavelet decomposition, multiresolution analysis is incorporatedinto the network. In order to identify the condition, the network is given the frequency sub-bands that were recovered from the input pictures. The network's goal is to determine whether the input image belongs to the Covid-19 class or not. The Advantage of the proposed system are that it could be the very first-of its kind, cost-efficient, and highly accurate application that provide complete and accurate covid - 19 diagnosis.

As the pandemic hits the world, our whole education system adapted to the new online system which has many perks and a few minor drawbacks. So to minimize all the flaws of the online education system we have made software that uses the Facial Recognition system for the attendance of the students. The approach is to use the AI Facial Recognition System, it uses the multiple layers and photos capturing technique to capture an accurate picture of a human face with help of Ai ML algorithms like AdaBoost, and Geometry-based algorithms. The given system is built on the TKINTER platform and has a PYTHON script and a SQL database to back it up. The system's algorithm compares images based on the values of the face in the database image and the real-time image taken in the system. To store the data of the users we have used pandas which feeds the input and output in Excel. Apart from this, we have some ML libraries which are pre-trained like haar cascade - for facial recognition and OpenCV. We needed to keep the software easy to use and small in size in which we succeeded by creating an admin authorization to add new students. We have put forward both our ideas and our abilities to create an “Automated Attendance System Based on Face Recognition.”

Recent years have seen a significant increase in scholarly interest in object detection because of its strong connection to video analysis and picture interpretation. Shallow trainable structures and handcrafted features facilitate conventional object recognition techniques. Intricate ensembles that integrate poor visual elements with high-level data from object detectors and scene classifiers reach an effectiveness threshold relatively quickly. In order to help with problems with conventional architectures, more powerful tools that can learn deeper, higher-level, and more semantic characteristics are becoming available as deep learning develops quickly. For instance, these models behave differently when it comes to network architecture, training strategy, and optimization methodology. In this paper, we evaluate studies on object recognition using deep learning. The authors of the study begin with a primer on deep learning and its principal methods, the convolutional neural network (CNN). The subject will then shift to a number of widely used generic object detection methods, as well as various improvements and practical strategies for improving detection generally. The topic of different common genealogical patterns for object recognition will next be covered, along with some tweaks and practical methods for enhancing detection. even more performance It also briefly explores a number of particular detection tasks, such as pedestrian identification, face detection, and recognition of remarkable items because they display a range of properties. Experimental analyses are also available to compare alternate strategies and get some useful results. The recommendations for further research in both the object and object-oriented fields cover a wide range of prospective directions and objectives.

Emojis are a type of emoticon that are often used in text messages sent across social media platforms. A contemporary manner of communication is characterized by the inclusion of both textual and graphical information inside the same message. Emoticons and avatars are both examples of non-verbal communication tools. These indicators have rapidly become an important component of a wide variety of activities, including online talking, product reviews, brand emotions, and many others. It also resulted in an increase in the amount of data science research devoted to narratives driven by emojis. It is now feasible, as a result of improvements in computer vision and deep learning, to identify human emotions based only on visual cues. In this deep learning project, we will classify human facial expressions in order to map and filter avatars or emojis that correspond. This project's goal is to make the talking world appear more vibrant; it is not intended to offer a solution to an issue that occurs in the real world. Emojis is a piece of software that makes it easier to create avatars and emojis.

This article has been withdrawn at the request of the Publisher. The Publisher apologizes for any inconvenience this may cause. The full Elsevier Policy on Article Withdrawal can be found at https://www.elsevier.com/about/policies/article-withdrawal.

Many people became more anxious of a sudden heart arrest lately. As smart wearable devices becoming more popular, an opportunity has become more open to deliver the Internet of Things (IoT) solution. Unfortunately, for patients recovering from sudden heart arrests, hospital survival rates are poor. The goal of this study is to create an IoT device that collects body area sensor (BAS) data with the intent of providing early warning about an imminent heart arrest. This study aims to establish a clinical surveillance method for cardiovascular patients who stay home (health care) and the advice on treatment (health care). The vital signs of patients (heart rate, blood pressure and cardiac rhythm) are tracked by means of sensors that report to the M2M server. The server would then immediately prescribe the drug using a case-based inference method called artificial intelligence (AI), which incorporates multiple hospital care reports and other sources. The stress analysis prediction model uses a machine learning algorithm, including a decision tree, a classifier K-neighbours and a support vector classifier. In a data ban known as the Blob storage scheme, the data of the patient are saved and secured and run by the SQL server. The new patient data can be interacted by processing via Kubernetes with current patients in data storage using the machine learning model framework. The comparable findings assess the stress thresholds and therefore the seriousness of the heart attack. The IoT sensor-based platform represents a safe framework to monitor, control and interpret patient information and to address the needs of specialty hospitals.

Heart disease is a leading factor in human health care without proper treatment due to huge volume of data processed in big data. Due to congestions occurrence in network Because of IoT communication is not as well defined. By analyzing routing defects and prediction accuracy to make proper data analysis to resolve the risk. The key risk factors for heart disease are obesity, smoking, alcohol consumption, and age factors. Prognostic systems are designed to reduce mortality based on IoT.. This paper provides large data analysis for the prognosis of coronary heart disease. With a large amount of development data in all fields, it is difficult to analyze, extract, manage, and configure data that are used in its large data technologies and tools. To propose anIntelligent Big Data Analytics Model(IBDAM) for Efficient Cardiac Disease Prediction with IoT Devices in WSN Using Fuzzy Rules. By the intent, a multi-level fuzzy rule generation estimated with Cardiac Disease Infection Transmission Analysis (CDITA)weight find the features which are carried to cardiac disease prediction. This feature is trained with an optimized recurrent neural network. The features are classified with labeled class based on the medical practitioner with the risk of evaluation. By predicting this class based on the risk, the diagnosis be carried out for premature treatment and early prediction. The proposed system produces high performance compared to other conventional systems.

Every human being in today’s age has high importance in a stable human life. With daily new technology is launched in a world loaded with innovative business sectors to tackle more effectively the prevalent challenges of the world. We ought to figure out if as many people as possible can be spared. The death rate of multiple non-communicable heart disease forms is rising steadily each year. Many non-communicable disorders are long-term, gradual and too serious that a patient’s situation is too critical to control. This leads to a sudden heart attack for most people or learning about their condition before it is too late. In this document we suggest a cardiovascular state prediction approach for IoT and Machine Learning, which will use IoT system (sensors) to capture the data needed from the human body and move it to the cloud where data is saved with user verification. The obtained information from the human body is then normalized to quantify and forecast the total condition before AI calculations are applied to them.

Our proposed system discusses the concept of a smart wearable device connected to their parent’s mobile phone for children and their parents respectively. In this project we propose that to let the system be divided into three parts, namely the safe, intermediate and danger zones. If the child is within the safe zone, then no buzzer is sounded whereas if the child is in the intermediate range a buzzer alert will be sounded. If the child crosses the ‘danger’ zone, the buzzer is sounded with an immediate notification sent to the parent. In case the child goes out of danger zone, a GPS module is attached that would help parent know the exact location of the child once he/she is outside the 100meters of radius from the parent. This project also has features to sense the child’s temperature and heartbeat along with notifying the child’s parent in case the child has an accident using the temperature, heartbeat and pressure sensors respectively. The RSSI is used for distance sensing whereas GSM is used for notification sending to the parent’s mobile phone.

In the recent years, with the advent of mass globalization, circuits that are being manufactured are not done end to end in the same roof. Some parts are manufactured by a company while others are being outsourced to others, sometimes even to different countries. Therefore, the security of the circuit is heavily compromised for quality and speed, thus forsaking the integrity. This is the reason why logic-locking was proposed and many methods and algorithms were tested and deployed to maintain the integrity of the circuit when it passes through IC integrator, foundry, assembly/test facilities, distributor before it reaches the end customer. While many of them have been successful, more and more attackers are finding a way to make them obsolete. One of the recent methods is by side channel analysis. This method is analysed and dealt with deeply. In this paper we demonstrate the design and implementation of a 128-bit Advanced Encryption Standard (AES) encryption algorithm and how an efficient side channel attack mollifies the strength of AES by deploying it on SASEBO-GIII FPGA board with the Xilinx Spartan and Kintex 7. The implementation has been tested successfully.