The COVID 19 pandemic is highly contagious disease is wreaking havoc on people's health and well-being around the world. Radiological imaging with chest radiography is one among the key screening procedure. This disease contaminates the respiratory system and impacts the alveoli, which are small air sacs in the lungs. Several artificial intelligence (AI)-based method to detect COVID-19 have been introduced. The recognition of disease patients using features and variation in chest radiography images was demonstrated using this model. In proposed paper presents a model, a deep convolutional neural network (CNN) with ResNet50 configuration, that really is freely-available and accessible to the common people for detecting this infection from chest radiography scans. The introduced model is capable of recognizing coronavirus diseases from CT scan images that identifies the real time condition of covid-19 patients. Furthermore, the database is capable of tracking detected patients and maintaining their database for increasing accuracy of the training model. The proposed model gives approximately 97% accuracy in determining the above-mentioned results related to covid-19 disease by employing the combination of adopted-CNN and ResNet50 algorithms.

[This retracts the article DOI: 10.1007/s11042-023-15640-2.].

Heart disease is on the rise in both young and old people in today's modern culture, which is marked by elevated levels of stress, anxiety, and depression. In addition to being common, these illnesses carry serious dangers to life. Timely detection and prevention of heart problems is essential for a successful recovery. After extensive study on prognosis and prevention, cardiac arrhythmia becomes a major issue impacting a significant proportion of the population. Every day in medical institutions, a large amount of data is generated by the electrocardiogram (ECG), which is the most economical way to diagnose cardiac arrhythmia. number of automated models for identifying cardiac arrhythmia have been developed recently using machine learning and deep learning methods. This paper performs a thorough analysis of these latest developments, rating their effectiveness according to certain criteria such as applied settings, deployed datasets, input data changes, approaches, and outcomes attained. Apart from these dimensions, the review also takes generalizability and interpretability into account. It also discusses the shortcomings of the current research and suggests directions for further development. The significance of resolving these issues is highlighted by this work in order to improve the efficiency and usefulness of automated cardiac arrhythmia detection algorithms.

Digital Authentication combined with IOT networks is one of the most innovative practices in innovations. IoT devices such as surveillance cameras have acquired real-time video data in the natural environment. Facial Biometrics for System authentication is considered to be a sophisticated technology used in various devices. The correct processing of digital information in a natural environment is one of the challenging tasks. Noise effect inducts random variations on the subject in the natural environment that may hinder the recognition efficiency of the model. The proposed system recognizes facial biometric features under the natural environment that may contain non-symmetrical randomness of variations. The system's objective is to mitigate the false recognition rates under adverse environmental conditions. It combines digital information with a person’s real-time environment. The detection of the facial region of interest form frames in a real-time dataset has been taken through the Viola-Jones algorithm. Feature extrication using video frames has been developed through Deep Reinforcement Learning (DRL) algorithm, which aims to generate binary trees containing feature vectors. Further, the system uses a convolutional neural network (CNN) model to establish the correlation of feature vectors belonging to the facial identity. The model's objective is to retain high recognition of facial biometric trait feature units under various randomness in a natural environment. The proposed system is also tested under various attacks to test the robustness of the proposed model. The model can secure an average accuracy of 98.85

Interpreting psychological events can be costly and quite complex. It is simple to translate such experiences into a person's spoken and nonverbal cues. The suggested model investigates a computer vision-based method for using an individual's audio signal to identify stressful psychological events. Different people's input speech signals are recorded and compared to the common questionnaire. A series of inquiries pertaining to the second stage of COVID-19 events are included in the questionnaire set. Through additional processing, these speech signals are converted into frequency components by means of the Fast Fourier transformation (FFT) method. A long short-term memory module processes each frequency component and produces temporal information from each frequency band. The features of speech signals are extracted into the temporal frames by this module. The VGG 16 algorithm is used to further classify each temporal frame into stress and un-stress classes. A classifier with 16 layers of architecture is called VGG 16. A feed-forward convolutional neural network called VGG 16 is used to divide the vast array of speech signal features into classes: stressed and unstressed. The proposed model attempts to recognize speech signals as stress indicators. A standard set of questionnaires with a series of interrogation-style questions has been used to develop the stress symptoms in an individual's mind. The audio signals generated by each person's responses are recorded and subsequently analyzed for stress and un-stress classes. The proposed model was able to identify stress in speech signals with 98% accuracy. The time and cost implications of the suggested model are relevant. Medical research is typically costly and time-consuming.LSTM; VGG 16; CNN model; data preprocessing; speech signal.

A highly contagious illness caused by the SARS-CoV-2 virus pandemic is proven to wreak havoc on people’s health and well-being all over the globe. Severe Acute Respiratory Syndrome Corona Virus 2 (SARS-CoV-2) is the source of COVID-19. Chest radiography is one of the most crucial databases for applying detection techniques. COVID-19 infects the respiratory system and replicates, affecting the alveoli as well. Conventional approaches, such as RT-PCR tests, rapid antigen tests, serological tests, etc., are generally used to detect COVID-19 and have proven costly and time-consuming. Several suggested artificial intelligence (AI)-based models for detecting COVID-19 in contaminated individuals use lung ultrasound images, voice patterns, chest sounds, etc. In this paper, we have proposed a lightweight CNN model with a ResNet50 modified configuration that has been used to identify COVID-19 cases using features and variations in the chest radiography image dataset. The empirical results and comparative analysis with the ResNet 101 model prove the lightweight nature of the proposed CNN model. A chest radiography dataset containing COVID-19-infected, normal, and pneumonia-infected images. The dataset comprised almost thousands of chest radiography images from patients from two open-access information standard repositories. The proposed lightweight model gives approximately 97% accuracy by combining the adopted CNN and ResNet50 algorithms.

With the increasing demand for intelligent transportation systems and autonomous vehicles, reliable and efficient road lane detection has become a crucial component for ensuring road safety and enhancing driving assistance technologies. This research presents an innovative approach to automatic road lane detection utilizing Convolutional Neural Networks (CNNs). The proposed system leverages the power of deep learning to analyze road images and accurately identify lane markings. A dataset comprising diverse road scenarios is used to train the CNN model, allowing it to learn complex features and patterns associated with different road conditions, lighting, and environments. The architecture of the CNN is designed to extract hierarchical representations, enabling robust and adaptive lane detection. The process involves preprocessing techniques to enhance image quality, followed by the CNN model's training and validation phases. The trained model demonstrates high accuracy and efficiency in detecting road lanes in real-time scenarios. Evaluation metrics such as precision, recall, and F1-score are employed to assess the performance of the developed system, ensuring its reliability and effectiveness. The proposed automatic road lane detection system exhibits promising results across various challenging conditions, including low-light situations, adverse weather, and diverse road geometries. The integration of this technology has the potential to significantly improve road safety, advance autonomous driving capabilities, and contribute to the overall enhancement of intelligent transportation systems. The findings of this research pave the way for future advancements in computer vision applications related to road infrastructure and vehicular safety.

The fact that brain tumors belong to the most fatal illnesses, prompt and precise detection techniques are necessary. Optimization of MRI scan is the very first approach in which pre-processing and post processing are used to identify the best image for the goals of the research. Consequently, a threshold was applied to divide up the MRI pictures by incorporating the mean grey level approach. Using Hara-lick's feature equations and the spatial gray-level dependency matrix, the second stage of statistical feature analysis involved the extraction of data (SGLD). As a result, the tumor was positioned correctly and the best features were chosen. In the third step, supervised learning and artificial intelligence techniques were used to create an automated tool that could classify the photos being evaluated as having a tumor or not. An effective network performance test produced 97% of the intended outcomes.

Psychological activities have various dimensions in which they correlate with their respective behavior generated by the human body. Understanding the relationship of psychological events with the help of external action units is one of the research subjects to explore various human behavior and their dependencies. Existing work applied various deep learning algorithms to outline the correlation between psychological activities with human emotions. The study of psychological analysis in the medical field is very time-consuming and costly. It requires constant monitoring of the patient for a period of time and various interrogation sessions to finalize the emotional severity of an individual. Few skilled specialists and the lack of medical knowledge of human emotions drive the need for computer vision approaches to emphasize emotion recognition, particularly in disorders. The proposed study specifically assesses the use of speech signals to identify psychological disorders in terms of stress characteristics and uses a deep learning model that incorporates feed-forward networks and long short-term memory (LSTM) to identify the degree of psychological disease. The study made use of a standard speech dataset that was gathered from a variety of patients using a standard questionnaire format. The survey was conducted throughout a few Indian states. Speech samples were taken from patients whose cortisol levels were higher than 10 %. To assess the relationship between speech and psychological activity, speech signals from each patient have been gathered. The spectrogram of the speech signal's Mel filter bank coefficients has been analyzed, and the characteristics that cause stress and those that don't have it have been further divided into categories. The suggested model classifies stress and non-stress features in 150 voice dataset subjects with an average accuracy of 98 %. The model is found to be robust for various applications such as preventing suicidal cases, improving decision-making in the diagnosis of depression patients, improves the overall mental healthcare system.

The COVID-19 started spreading from China to other parts of the world in 2019. COVID-19 shook the health care system of the countries to a great extent even developed countries found it difficult to tackle the spread and provide necessary services in the health care sector. It has shown its effects in the world at an obstreperous rate. During the mid-2020, it took millions of lives, this devastating effect has made it come into the limelight of rapidly growing technology. All the unfortunate results of the COVID-19 virus have motivated the Deep Learning model, which will help doctors and lab assistants in categorising COVID-19 using X-rays imaging. To date, we are encountering further variants like omicron (recently). Some already existing convolutional neural networks (CNN) have shown promising results in detecting infected patients from X-Ray.

Biometric applications have massive demand in today's era. The areas of applications are mostly linked with the security of the system. Biometric features are regarded as the primary resource for security purposes due to their own distinctiveness and non-volatile essence. System authentication using biometrics is considered to be a sophisticated technology. Noise effect inducts variation in the biometric subject that causes an adverse impact on establishing the recognition. The proposed model supported the development of an effective method for performing facial biometric feature recognition. The model's goal is to reduce the number of false approvals and refusals. The proposed algorithm has been applied over a video dataset containing surveillance video frames that captures facial subjects dynamically. The first step is pre-processing of the video frames that have been carried out in the proposed model. Then, the Viola-Jones algorithm was applied to detect the facial subjects in the video frames. Feature extraction from the facial subject has been accomplished by applying a deep reinforcement learning algorithm. Further, the proposed model applied a convolutional neural network (CNN) algorithm to perform feature recognition of facial identity accurately. The proposed technique aims to maintain a huge recognition rate of dynamic facial subjects under various unprecedented noise variations. In the classification algorithm, the recognition accuracy is found to be 98.85%.

Brain tumors are one of the deadliest diseases and require quick and accurate methods of detection. Finding the optimum image for research goals is the first step in optimizing MRI images for pre- and post-processing. As a result, the mean grey-level approach was used to segment the MRI images using a threshold. In the second stage of statistical feature analysis, the picture features were extracted using the spatial gray-level dependency matrix using Hara lick's feature equations. As a consequence, the best features were picked and the tumor was placed in the appropriate location. In the third stage, an automated tool for classifying the photographs under assessment as having a tumor or not was created utilizing supervised learning and an artificially intelligent methodology. An efficient test of the network's performance yielded 97

Biometric applications have massive demand in today’s era. The areas of applications are mostly linked with the security of the system. Biometric features are regarded as the primary resource for security purposes due to their own distinctiveness and non-volatile essence. System authentication using biometrics is considered to be a sophisticated technology. Noise effect inducts variation in the biometric subject that causes an adverse impact on establishing the recognition. The proposed model supported the development of an effective method for performing facial biometric feature recognition. The model's goal is to reduce the number of false approvals and refusals. The proposed algorithm has been applied over a video dataset containing surveillance video frames that capture facial subjects dynamically. The first step is the pre-processing of the video frames that have been carried out in the proposed model. Then, the Viola-Jones algorithm was applied to detect the facial subjects in the video frames. Feature extraction from the facial subject has been accomplished by applying a deep reinforcement learning algorithm. Further, the proposed model applied a convolutional neural network (CNN) algorithm to perform feature recognition of facial identity accurately. The proposed technique aims to maintain a huge recognition rate of dynamic facial subjects under various unprecedented noise variations. In the classification algorithm, the recognition accuracy is found to be 98.85%

Identifying brain tumors and improving patient care are the goals of this project. Brain malignant tumors, which are aberrant cell growths, are known as tumors. Frequently, infectious brain tissues are caught by CT and MRI scans. There are many other methods that are used for the detection of brain tumors and some of them are positive charges imaging and cerebral X-ray photography of blood or lymph vessels and tests at the molecular level. So, this paper will use different MRI images to detect ailment cause like tumors. The main purpose of this research paper is to 1) identify irregular sample images and 2) find the tumor area. The abnormal sections of images will forecast the levels of tumors so that proper treatment can be done. Deep learning is used to find abnormal regions from sample images. This paper will use VGG-16 to segment the abnormal part. The density of the infected region is defined by the number of pixels with malignancy.

Medicines have always been of the utmost importance in every era due to their curing properties. Now-a-days, medicines for almost every disease are available. Also, different kinds of medicinal systems have come into existence. Despite the present era, in the ancient era, there existed only one medicinal system, known as Ayurveda that is considered the backbone of medicinal systems. The medicines are prepared from medicinal plants. Now-a-days, many countries have moved to Ayurveda. Medicinal plants are harvested in a similar manner as food crops are harvested in agricultural fields. Diseases in plants reduce the quality and quantity of the product. Also, the medicines would not be useful if prepared using diseased plants. Thus, monitoring health is a must. Manual inspection is a tiring task with a huge loss of budget and time, and the loss increases with the size of the agricultural field. Thus, image processing techniques have proven to be beneficial for detecting, identifying, and classifying diseases in medicinal plants, as they reduce the need for tiresome field inspection and also save time and money. Diseases can be detected as early as when they start appearing on the surface of the plants, thus helping in the taking of appropriate preventive measures to stop the growth of the disease and even prevent it from occurring in the future. Detection and classification of diseases comprise steps of image processing. The different detection techniques are described. Also, a new technique is proposed for identifying and classifying diseases in medicinal plants.

In a biometric-based security system, rather than depending on the system configuration itself, failure rate also relies upon feature extraction and its related statistics. In this paper, a significant approach is being presented to minimize the failure rate and maintain high recognition accuracy and uniformity for non-symmetrical feature points. This work contributes a detailed analysis of stable parameters of captured biometric feature points by using a flexible learning model named as adopted Artificial Neural Network (ANN). The paper also discusses a comparative study of different global and local methods of Histogram of an Equivalent Pattern (HEP) technique for facial feature detection and extraction. The HEP, is further classified by using adopted ANN model, which depends on partitioning the feature area of a predefined image. This task has been accomplished by providing the appropriate definition of local and global functions based on pixel intensities. The literature available for face detection shows many shortcomings such as false acceptance and rejection rates. Among all defined global and local techniques, this paper primarily endorsed an adopted method of Improved Local Binary Pattern (ILBP) which works on local pixel values of a facial image for feature extraction. The classification and recognition task are performed by adopted ANN for various defined global and local features. The paper also derives a detailed comparison with the other existing techniques. As a result, the proposed ILBP technique ensures the consistency of acceptable results in unpredictable variations in the dataset.

The security of a system through biometric features is one of the acceptable trends to which every researcher recommends. Such biometric features contain features of iris, fingerprint, face, etc. Today, these security traits are widely exploited by an imposter to trace unlicensed access to the system. So it demands to secure these features since foremost alarming challenges are raises when these trustworthy features are compromised. This paper endorses the design of zero-bit watermarking in which a user’s unique id is integrated with his facial features in the multi-transform domain. The user’s unique ID is used to establish true authentication of the host facial features and also to recognize the face without any segmentation techniques. To make this effort tangible, Singular values are calculated of some particular frequency coefficients of the host image that are generated using FFT transformation in which a number of appropriate regions of the host image are selected based on their frequency of information. These Singular values of host facial image are calculated using SVD in which watermark (image) user’s unique ID is integrated correctly while maintaining the equilibrium among imperceptibility, robustness, and payload. The resultant watermarked image is tested against various image processing attacks and satisfying results are conducted to assure robustness of the model. Thus, the model is indeed satisfying security of the host biometric of the user.

Biometric security has long been a trending zone that satisfies the need for a significant level of security and control. Among all the existing technologies, face detection is one of the most utilized and adjusted innovations. The identification failure of a user's identity is a big concern. In this chapter, a novel approach for biometric recognition has been introduced in which the application of ILBP (Improved Local Binary Pattern) for facial feature detection is discussed which generates improved features for the facial pattern. It allows only an authenticated user to access a system, which is better than previous algorithms. Previous research for face detection shows many demerits in terms of false acceptance and rejection rates. In this paper, the extraction of Facial features is done from static and dynamic frames using the Haar cascade algorithm. Then, the ILBP method which works on local pixel values of an image is applied for feature extraction, and finally, the SVM (support vector machine) is used for classification of those features. The objective of this paper is to provide the best recognition results from images that are taken randomly and may possess noise. This paper achieved an accuracy of 97.90% for correct recognition and with less time complexity. It can be used in crime investigation, security cameras, digital forensics, etc.

Unlicensed access to digital audio is found to be very frequent today. Copyright and ownership issues are very common. Obsolete schemes are looking to fail to preserve the ownership of digital audio. A watermark data validates the correct belongingness of an audio file. Inculcation of any data into an audio signal is not always tolerated by audio. In this work, a robust method is proposed based on a hybrid decomposition technique in which discrete wavelet transform (DWT), discrete cosine transform (DCT), and singular value decomposition (SVD) applied to successfully perform watermarking in audio signals. A watermark image of size 16 by 16 pixels is used to inculcate in the digital audio of sampling rate 44.1 kHz. The watermark image is first undergone into a cyclic encoding process in which watermark bits are encoded using the redundant bit. Then the bits are further scrambled using Arnold’s cat map. After the robust encryption, the encrypted watermark bits are embedded into the host audio using hybrid decomposition. The reverse process is applied to extricate the watermark bits. The watermarked audio is tested under various signal processing attacks and the quality of the extracted watermark image is checked using standard parameters. The quality assessment of the watermark image is found satisfactory that declares the robustness of the scheme.