Efficient waste management is crucial in today's environmental landscape, necessitating comprehensive approaches involving recycling, landfill practices, and cutting-edge technological integration. The proposed approach introduces a sophisticated waste management system, harnessing dual or twofold convolutional neural networks (D-CNN or TF-CNN) and a histogram density segmentation (HDS) algorithm. This intelligent system equips users with the means to enact essential safety protocols while handling waste materials. Notably, this research presents groundbreaking contributions: Firstly, a geometrically designed smart trash box, incorporating ultrasonic and load measurement sensors controlled by a microcontroller, aimed at optimizing waste containment and collection. Secondly, an intelligent method leverages deep learning for the precise classification of digestible and indigestible waste through image processing. Lastly, a cutting-edge real-time waste monitoring system, employing short-range Bluetooth and long-range IoT technology through a dedicated Android application was proposed.

The importance of Machine Learning in predicting Air Quality Index (AQI) is increasing day by day. This aids in informed health protection decisions amid air pollution's impacts. The factors that affect air quality and the machine learning model parameters that yield good prediction accuracy are not well covered in the existing literature. This study provides a comprehensive comparison analysis that estimates air quality using multiple regression models. To forecast the Air Quality Index, we utilize data from two distinct places in Chennai City. Manali, an industrial area, and Velachery, a residential suburb, are the two. For the purpose of assisting us or the government authorities in making decisions, the study also focuses on the factors that influence air quality. The main goal is to determine which model performs best by evaluating a variety of factors, including R2 measurements and Mean Absolute Error.

Alzheimer’s Disease (AD) is a degenerative, chronic condition of the brain for which there is now no effective treatment. However, there are medications that can slow its development. In order to stop and control the development of AD, earlier diagnosis of the disease is quintessential. Our proposed method’s primary objective is to establish a comprehensive model for the prior detection of Alzheimer’s disease and the categorization of distinct AD stages. This work employs a deep learning methodology, especially CNN. The proposed approach makes use of well-known models that have already been trained to classify medical images, like the EfficentNetB7 model, by applying the transfer learning principle. In order to achieve greater accuracy, convolutional neural networks (CNNs) are frequently scaled up as new resources become available at a fixed cost throughout the construction phase. A compound coefficient is used by the CNN architecture and scaling approach, which is the foundation of the pre-trained EfficientNetB7 model, to scale the dimensions equally. This proposed EfficientNetB7 model is quicker, easier, and more effective than other pre-trained models like VGG19 and InceptionV3. The proposed model includes simple structures that have memory requirements, provide manageable time, overfitting, and low computational complexity as well as training and inference speeds. The Alzheimer’s disease Neuroimaging Initiative (ADNI) dataset was employed for a comprehensive assessment of the method proposed, utilizing well-known performance metrics including sensitivity, specificity, and accuracy. The findings revealed that the improvised results achieved the accuracy metric when compared to existing methods. The EfficientNetB7 model has been enhanced, and this model achieves a sensitivity of 98.08

Drowsiness among drivers is one of the leading causes of road accidents and deaths. In this research, we propose a new way to detect driver drowsiness using MobileNet architecture with transfer learning. The goal of the proposed system is to improve the accuracy and effectiveness of Drowsiness Detection in vehicular environments.The first step in the research is to train MobileNet using an existing image classification dataset. Transfer learning is then used to adjust the model for the specific drowsiness detection task. Transfer learning improves the model's capacity to generalize from a limited set of labeled data. Combining MobileNet's deep learning capabilities with transfer learning allows us to leverage the advantages of both techniques, balancing computational efficiency with accuracy. Haar Cascade Classifiers can be used to analyze facial features, including eye tracking, facial expressions, etc. This classic computer vision technique adds an extra layer of detection, increasing the robustness of the overall drowsiness detection solution. Combining the deep learning capabilities of MobileNet with the real-time efficiency of haar Cascade Classifiers provides a complete solution for drowsiness detection in real-time. The proposed system with a validation accuracy of 86% has the potential to be incorporated into advanced driver assistance systems (ADAS) systems to improve road safety by alerting drivers when they are drowsy. This research opens the door to intelligent adaptive systems that could play an important role in reducing the number of drowsiness-related accidents and saving lives.

Driving a car is a challenging task that requires your full attention. Any action that causes the motorist to lose focus on the road constitutes distracted driving.1.35 million people perish in automobile accidents occur annually. Approximately one out of every six car accidents are caused by distracted driving, according to the National Highway Traffic Safety Administration. We want to build a system that can identify distracted drivers and alert them. To accomplish these goals, we gathered a wide range of data that included instances of distracted driving activities, such as texting, making phone calls, eating, and adjusting the radio. To guarantee consistency and suitability for model training, the dataset underwent preparation and was assigned tags with matching class labels. After that, we created a convolutional neural network (CNN) architecture that was enhanced for image classification tasks by using transfer learning strategies. The outcomes of the study show how well the suggested method works in real time for recognizing distracted driving behaviors. Over 90% classification accuracy was achieved by the trained model on a held-out test dataset, exceeding baseline techniques and proving its usefulness in real-world situations. Additionally, we put in place an intervention mechanism that will promptly notify drivers when distracted driving behaviors are identified, raising awareness and supporting safer driving practices. All things considered, this research advances the development of devices meant to lower the prevalence of distracted driving and improve traffic safety. We offer a viable framework for tackling this important problem and possibly saving lives along the road by utilizing deep learning methods for real-time detection and intervention.

Using a cutting-edge neural network framework, “PulseSync BP” estimates blood pressure without contact. Using photoplethysmogram (PPG) signals and other physiological data, this novel model uses advanced signal processing. The suggested framework changes blood pressure monitoring by eliminating cuff-based measures. PulseSync BP's RNN architecture is carefully designed for optimal prediction accuracy. The algorithm effectively estimates blood pressure through extensive experimentation, demonstrating its potential for continuous healthcare applications. This framework's capacity to use various physiological cues shows its adaptability and versatility. PulseSync BP allows real-time, non-intrusive blood pressure monitoring, advancing individualized health tracking solutions. Cuffless blood pressure estimate is advanced here, demonstrating how artificial intelligence could improve healthcare accessibility and patient-centric monitoring. PulseSync BP and other revolutionary technologies can transform blood pressure monitoring in healthcare.

Prediction of invoice payment date through machine learning in the B2B marketing is an important factor that affect the business dealings between the companies and changes the complete direction of the business. If the seller company finds the predicated payment date of the buyer company is getting highly delayed from the due date then seller company may not sell that product to that company. So, in this way, payment date prediction plays a very important role in B2B marketing. In this study, we explore how machine learning (ML) can be used to develop models for predicting whether newly created bills will be paid, enabling customized collection activities specific to each invoice or customer. Our models can accurately forecast whether or not a bill will be paid on time and also give estimates of how much time will be lost. Our methods are demonstrated using real-world transaction data from several firms. Finally, simulation results compared with other state-of-the-art approaches.

Crowd density estimation is a significant research area in artificial intelligence applications as it is an effective tool for crowd monitoring, control, and behavior comprehension. Unstructured big data includes a substantial contribution from surveillance videos. Manual surveillance appears to be time-consuming and tedious. As a result, many CNN oriented detecting people and estimating population models were published in recent years. The enormous growth in crowd counting methods observed during the last several period is typically due to significant advancements in CNNs and datasets. Large crowd datasets and the recent advent of deep learning have made it possible for most crowd counting techniques to be remarkably successful. We have reviewed current articles on deep learning-based algorithms to recognize and categorize the regions with high crowd density in surveillance scenes to estimate the crowd density for the provided crowd scene image. This study analyzes current efforts and briefly discusses regression, detection and conventional-based density estimation methods. On video surveillance systems, several techniques mainly focus on accurately identifying and classifying crowd density. Considering a focus on using camera footage to estimate crowd volume, this overview intends to summarize studies that are crucial to the wider region of video analysis using deep learning methods. This work presents a thorough assessment along with an examination of the best current developments and efficiency improvements for assessing the density of crowds in observation settings. To ascertain the outstanding effectiveness and accuracy of the current modern techniques for deep learning population density estimation, their effectiveness is evaluated.

Given the increase in population, vehicle tracking is no longer practical. Both time and resources are wasted in doing it. The enormous daily growth in the automotive industry has made tracking individual automobiles an extremely challenging undertaking. This research suggests a system for automatically tracking moving cars using roadside security cameras. License plate recognition systems are used in toll collection, parking fee, and residential entry control in contemporary smart cities. In addition to being helpful in people's daily lives, these electronic technologies also give management access to secure and effective services. The suggested technique now includes a useful method for identifying Indian license plates on cars. The suggested technique can work with number plates that are obtrusive, dimly lit, cross-angled, and have unusual fonts. The effective deep learning-based ALPR (Automatic License Plate Recognition) model presented in this study uses character segmentation and a CNN-based recognition model. The experimental finding yields a 94.94% accuracy percentage for the f1 score.

Alzheimer’s disease is a neurological condition that gradually reduces brain size and destroys brain neurons. The most common varieties of dementia, Alzheimer’s disease limits a person’s ability to operate independently and is described by a steady deterioration in mental, cognitive and social abilities. Due to the severity of moderate cognitive impairment, AD diagnosis is frequently challenging at an early point. Nevertheless, treatment is likely to be successful at this stage. This raised concerns about the early diagnosis and treatment of AD. Deep learning approaches are used, including fastai and InceptionV3 to identify the most accurate markers for predicting AD.

Voting plays a very important role in electing a candidate in the government and private organizations these days. Voting system’s integrity and privacy are the main challenges in developing a voting system. The person who is voting a particular party should not be known by everyone which is a concern to privacy of the voter and the voting system should give correct results which shows the integrity of the voting system. One single vote may be responsible to change the entire result of the polling. So, voting system should be designed in such a way that each vote should be given its own importance and that should not get tampered by the hackers. So to solve this problem we are developing a full stack web application on voting system which does not compromises with integrity, privacy and security. This project has been developed using a server-side scripting language PHP. As the name server-side suggests the code which has been written cannot be accessible to the client thus it ensures privacy and security.

This chapter highlights the difficulties encountered during the model's training as well as the proposed approach for the model that has been used to address the issue of Alzheimer's disease detection systems. Federated learning is a model that has lately gained popularity because it shows great potential for learning from private information that is dispersed. The suggested technique is thoroughly examined using metrics like precision, recall, and accuracy on the Alzheimer's Dataset, which comprises 6400 MRI pictures. The system for detecting Alzheimer's disease is concluded in the fourth section, along with plans for improvements. The pictures from each class are as follows: The dataset was thoroughly examined using several different channels, and it was discovered that neither multiple images from the same patient nor repeated MRI scans of the patients were present in the dataset.

Abstract: One of the crucial areas of research in the field of advanced driver assistance systems (ADAS) is the detection and recognition of traffic signals in a real-time environment. These are specifically developed to work in real-time to improve road safety by informing the driver of various traffic signals such as speed limits, priorities, restrictions, and so on. This research paper proposes a traffic sign identification system on an Indian dataset utilizing the YOLOv5 model. This study suggests a method for detecting a particular set of 10 traffic signs. You Only Look Once (YOLO) v5 is the algorithm used to detect traffic signs, and the model parameters are trained on train sets obtained from the recently constructed dataset. The remaining images from the dataset are utilized to create a test set. When tested on the test set made from the suggested dataset, the proposed approach for detecting a particular set of traffic signs performs admirably

Mycobacterium tuberculosis causes tuberculosis (TB), a bacterial illness. Although the germs are most typically found in the lungs, they can affect other sections of the body as well. Tuberculosis is one of the primary causes of mortality in both developed and developing nations, necessitating worldwide attention. Even though TB may be prevented in the majority of instances if discovered and treated early, the number of deaths caused by the disease is quite high. There has been a significant increase in interest and research activity in TB detection in recent years. The new advancement in the field of AI Technology may be able to assist them in overcoming these development gaps. Computer-Aided Detection and Diagnosis (CADD) aids in the diagnosis of diseases by analysing symptoms and X-ray images of patients. Many solutions are currently being developed to improve the effectiveness of TB diagnosis classification using AI and DL approaches. Although a variety of TB detection techniques have been developed, there is no commonly acknowledged method. The purpose of this study is to give a survey on Tuberculosis Detection. It also emphasises the difficulty and complexity of the Tuberculosis Detection System's design.

Tuberculosis (TB) is still one of the most serious health issues today with a high fatality rate. While attempts are being made to make primary diagnosis more reliable and accessible in places with high tuberculosis rates, Chest X-rays has become a popular source. However, specialist radiologists are required for the screening process, which could be a challenge in developing countries. For early diagnosis of tuberculosis utilizing CXR images, a complete automatic system of tuberculosis detection can decrease the need for trained staff. Various deep learning and machine learning technologies have been introduced in recent years for examining digital chest radiographs for TB-related variances with the goal of reducing inter-class reader variability and reproducibility, as well as providing radiologic services in areas where radiologists are not available. Tuberculosis is sometimes misclassified as other conditions with similar radiographic patterns as a result of CXR images, resulting in inefficient therapy. The current approach, however, is limited to Computer-Aided Detection (CAD), which has only been evaluated with non-deep learning models. Deep neural networks open potentially new avenues for tuberculosis treatment. There are no peer-reviewed studies comparing the effectiveness of various deep learning systems in detecting TB anomalies, and none compare multiple deep learning systems with human readers. In this paper, the aim of the proposed method is to develop an efficient tuberculosis detection system based on stochastic learning with artificial neural network (ANN) model by random variations using Chest X-ray images. This approach can able to incorporate random functions into the network, either by assigning stochastic transfer functions to the network or by assigning stochastic weights to the network. This proposed method is to learn features from CXR images and optimize the parameters of an ANN model by randomly mixing the training dataset before each iteration, resulting in varied ordering of model parameter updates. Furthermore, in a neural network, model weights are frequently initialized at a random beginning point. By focusing on randomness functions with optimization, the proposed technique achieved great accuracy. The motivation of the proposed method is to detect abnormalities in CXR with the different levels of complexity of TB by strong or weak evidence with different deep geometric contexts such as shape, size, cavitation, and density. ANN’s primary benefit is extracting hidden linear and non-linear inter-relationships of high-dimensional and complex data. The proposed method was systematically tested with the Shenzhen and Montgomery datasets using metrics such as sensitivity, specificity, and accuracy, and it was discovered that the proposed method attained better accuracy when compared to state-of-the-art methods. The proposed method shows an improved efficiency with sensitivity of 96.12%, specificity of 98.01%, accuracy 98.45% and F-Score 95.88% respectively.

Internet of things (IoT) devices are susceptible to numerous safety intimidations because of their limited features and abilities. In IoT, the wireless communications are generally shown intermittently amongst the power-limited nodes. IoT devices could certainly be seized, for instance, causing a node repetition attack. The wireless sensor network (WSN) should have an effectual, precise and reckless recognition mechanism that can perceive the occurrence of jammers and replicated nodes in the network. In this paper, support vector machine (SVM)-based cloning and jamming attack detection technique for IoT WSN is proposed. In this technique, the base station (BS) classifies nodes as cloned or normal by checking the distance measurements from the IoT devices. Simulation results have shown that the proposed SVM clone achieves high detection accuracy with reduced false positive rate and energy consumption.

Water scarcity is the major problem presently being faced globally; it is to be managed in an efficient manner. The water management procedure is one of the techniques to condense necessary water. The objective of water management system is that water supply agency to collect, distribute quality water without any delay and scarcity. An intelligent system is necessary for efficient production, collection and distribution. The proposed intelligent system consists of genetic operations with fitness value and neural network for training. The fitness function is used to make new intelligent members from existing population of water resources for water collection and distribution. The system is applicable for prediction about water consumption, distribution using decision-making algorithms to increase optimization performance by calculation of objective function of various population types. The regression performance of proposed intelligent system is calculated and compared with other algorithms.

Early Detection of Plant Leaf Detection is a major necessity in a growing agricultural economy like India. Not only as an agricultural economy but also with a large amount of population to feed, it is necessary that leaf diseases in plants are detected at a very early stage and predictive mechanisms to be adopted to make them safe and avoid losses to the agri-based economy. This paper proposes to identify the Tomato Plant Leaf disease using image processing techniques based on Image segmentation, clustering, and open-source algorithms, thus all contributing to a reliable, safe, and accurate system of leaf disease with the specialization to Tomato Plants.

Child security is the foremost common issue emerging around the world. There are numerous issues to youngster security and this work primarily manages kid security from the dangers like missing, abducts. The Technical point of this task is to have an ordinary correspondence between the kid and parent through the gadget which helps in finding the area, pulse and temperature of the kid utilizing the gadget empowered with the pulse sensor, temperature sensor and GPS tracker. This gadget empowers association between the youngster and parent through the WIFI module cooperation utilizing IoT. The parent can get to the kid data intermittently by interfacing through this gadget. This makes guardians defend youngsters even in their nonattendance. The data is stored into a cloud permanently to keep the track record of old data of the children for further reference. The sensors are activated automatically when they are subjective to the miscellaneous activities.